Descriptive

Column

About the study

Objective: The study aims to explore the awareness and use of the list of diagnosed medical conditions in Part C eligibility determination, from the perspective of the Early Intervention and Early Childhood Special Education (EI/ECSE), and healthcare professionals in the state of Oregon.

Methods: The embedded mixed-method design was adopted where the qualitative (interviews) data were nested within the quantitative (an electronic survey). Data from the survey and semi-structured interviews were collected simultaneously between November 2020 to January 2021. Responses were analyzed using descriptive statistics and thematic coding.

Results: A total of 90 EI/ECSE and healthcare professionals responded to the electronic survey and 18 professionals attended the semi-structured interviews. The mixed-method analysis indicates that 74% of professionals are aware of the eligible medical conditions, however, when they were shown the Oregon list of medical conditions, they reported not to have seen the list before. The majority of professionals (77%) reported they seek help from colleagues to decide EI referrals and eligibility evaluation instead of using the list. As an urgent response to the covid-19 pandemic, professionals (94%) shifted to an online platform for evaluations, assessments, and other services. Professionals reported a severe decrease in the number of referrals (48%) and delay in EI eligibility determination (62%) for children with diagnosed medical conditions. Professionals greatly agreed in their replies that the current list does not contain adequate information (65%) and they do not refer to it for eligibility determination. The three most agreed recommendations are to have a more comprehensive, detailed list(86%) especially during covid-19 (79%), widely circulated and embedded within the current system such as in the universal referral form and the evaluation packet in EI programs.

Conclusion: Lack of awareness, access, and inconsistent use of the list of diagnosed medical conditions that give children direct access to Part C EI/ECSE services can lead to an imbalance in number and timely receipt of Part C EI services in Oregon. A comprehensive list in a user-friendly language is needed. The list needs to be embedded in the existing systems for its long-term sustainable use and seamless cross-sectoral referral and evaluation.

Protocol: 09032020.003

Descriptive analysis

Participants N Percent
Profession
EI/ECSE Personnel 54 60
Healthcare Practitioner 36 40
Job Status
Full-time 73 81
Part-time 17 19
Role
Making Referrals to EI and Providing Med. Diagnosis 49 54
Part C EI/ECSE Eligibility Evaluations 41 46
Years of Service
Above 15 years 38 42
Less than 15 years 52 58

Column

Chart A: Survey participants across Oregon

Participants represented all 36 counties in Oregon. “Multiple” includes participants working in more than one county.

Awareness

Column

Survey participants fully completed this section.

90

Column

Questions in the survey (Awareness)

Questions N
Q1: Are you aware of the medical conditions for children that have a high probability of significant developmental delay listed on Oregon State Plan Policy guidance? 90
Q2: Are you aware that there are medical conditions for children that may qualify for EI eligibility evaluation and services that are not listed on Oregon State Plan Policy guidance? 90
Q3: Are you aware that there are medical conditions for which minimum criteria need to be met to be eligible for EI services and a child’s disability does not need to be presently adversely affecting the child’s development? 90
Q4: Are you aware of the medical documents and reports that contribute to the written documentation of eligibility for EI services based on the medical condition? 90

Analysis: Fisher’s exact test

Ques EI/ECSE Personnel Vs Healthcare Practitioner Full-time Vs Part-time Above 15 years of service Vs Less than 15 years Part C EI/ECSE eligibility evaluations Vs Making referrals and providing diagnosis
Q1:Aware of Med. Condition 0.0001*** 0.01*** 0.226 5.665
Q2:Elig Med. Cond. Not Listed 0.001*** 0.588 0.053 0.055
Q3:Med. Cond. Presently Not Affecting 0.001*** 0.399 0.651 5.665
Q4:Written Documentation of Eligiblity 2.295 0.548 0.101 6.305
Note:
***Fisher’s exact test p<0.05 at 95% confidence interval

Row

All participants

By profession

pdf 
  2

By Job status

Years of service

By professional role

In-practice

Column

Survey participants fully completed this section.

90

Column

Questions in the survey (In-Practice)

Questions N
Q: Have you ever experienced a delay in the EI eligibility determination process due to a lack of information or clarity on a child’s medical condition? 90
Q: Have you come across children with a diagnosed medical condition who could not access EI eligibility evaluation and other services on time? 90
Have you ever asked another professional for information on medical conditions to help evaluate a child’s EI eligibility? 90
Q: Have you ever searched for information on medical conditions to make EI eligibility determination? 90
Q: Have you ever referred a child directly for IFSP development based on a diagnosed medical condition, without doing the preliminary step of eligibility evaluation? 90

Row

By all participants

pdf 
  2

By profession

pdf 
  2

By job status

By years of service

By professional role

During Covid-19

Column

Total number of respondents fully completed this section of the survey.

87

total number of EI/ECSE professionals fully completed this section of the survey.

37

Column

Questions in the survey(COVID-19: For all participants)

Questions N
Q: Have you experienced a decrease in the number of children and families in need of referral for EI eligibility evaluation or medical diagnosis during Covid-19? 85
Q: Have you experienced a delay in EI eligibility evaluation and assessment for any child diagnosed with a medical condition during Covid-19? 85
Q: Are you conducting EI eligibility evaluations remotely through phone and videoconferencing for children during Covid-19? 85

Questions in the survey(COVID-19: For EI/ECSE participants only)

Questions N
Q: Are you following the 45-day timeline to do EI eligibility evaluation, assessments, and IFSP development for children during Covid-19? 37
Q: Have you enrolled children with a diagnosed medical condition with interim IFSP where EI eligibility evaluation was not possible during Covid-19? 37
Q: Have you met children or families in-person or made home visits for EI eligibility during Covid-19? 37
Q: Have you skipped EI eligibility evaluation for children with a diagnosed medical condition who have written documentation of eligibility from their physician during Covid-19? 37
Q: Are you using only tools that can be administered by parents to do EI evaluations for children during Covid-19? 37

Row

By all participants

pdf 
  2

By EI/ECSE Professionals

pdf 
  2

Views

Column

Total number of respondents fully completed this section of the survey.

81

Column

Questions N
Q: In your opinion, a more detailed and comprehensive medical conditions list would have been helpful during Covid-19? 81
Q: In your opinion, would a more detailed and comprehensive list of medical conditions help to decide when EI eligibility evaluations can be skipped and direct referral be made for EI assessment? 81
Q: In your opinion, is having a more detailed and comprehensive medical conditions list for EI eligibility determination in Oregon needed? 81
Q: Do you think that information on medical conditions listed in the Oregon State Plan policy guidance is adequate for deciding to carry out an EI eligibility evaluation? 81

Row

By All Participants

# A tibble: 16 × 5
# Groups:   position, response [4]
   position                response information                    count percent
   <chr>                   <chr>    <chr>                          <int>   <dbl>
 1 EI/ECSE Personnel       No       "Adequate information\n is gi…    27      33
 2 EI/ECSE Personnel       No       "Detailed list\nis needed ?"       7       9
 3 EI/ECSE Personnel       No       "Detailed list\n would be hel…     5       6
 4 EI/ECSE Personnel       No       "Detailed list during\n Covid…    13      16
 5 EI/ECSE Personnel       Yes      "Adequate information\n is gi…    21      26
 6 EI/ECSE Personnel       Yes      "Detailed list\nis needed ?"      41      51
 7 EI/ECSE Personnel       Yes      "Detailed list\n would be hel…    43      53
 8 EI/ECSE Personnel       Yes      "Detailed list during\n Covid…    35      43
 9 Healthcare Practitioner No       "Adequate information\n is gi…    26      32
10 Healthcare Practitioner No       "Detailed list\nis needed ?"       4       5
11 Healthcare Practitioner No       "Detailed list\n would be hel…     3       4
12 Healthcare Practitioner No       "Detailed list during\n Covid…     4       5
13 Healthcare Practitioner Yes      "Adequate information\n is gi…     7       9
14 Healthcare Practitioner Yes      "Detailed list\nis needed ?"      29      36
15 Healthcare Practitioner Yes      "Detailed list\n would be hel…    30      37
16 Healthcare Practitioner Yes      "Detailed list during\n Covid…    29      36

By Profession

pdf 
  2

INTERVIEWS

Column

Total number of interview participants

18

EI/ECSE professionals

11

Pediatricians

4

Physical Therapists

2

Speech and Language Pathologist

1

Row

Theme 1: Medical Conditions List

Theme 2: In-Practice

Theme 3: During Covid-19 Pandemic

Theme 4: Recommendations

Results

Row

Research Question1:

To what extent are EI professionals aware of the categories of medical conditions that are listed in the Oregon State Plan policy guidance? (AWARENESS)

Overall analysis showed that 74% are aware of the medical conditions listed in the Oregon State Plan Policy guidance, 72% are aware of the medical documents and reports that constitute written documentation of eligibility, 68% know that there are medical conditions that may not be presently adversely affecting child’s development, and 56% are aware that there are medical conditions eligible for Part C services but may not be listed in the policy guidance. When results were analyzed by profession, significant differences were found in the awareness of healthcare and EI/ECSE professionals where 21% of healthcare professionals are aware and 19% are not aware of the medical conditions list. Among healthcare professionals, 18% are aware and 22% are not aware of the written documentation of Part C eligibility, 19% are aware and 22% are not aware that there are medical conditions that may not be presently affecting the child’s development, and only 13% are aware and 27% are not aware that there are medical conditions that may be eligible for Part C services but are not listed in the policy documents. No significant differences were found based on job status (full-time vs part-time), years of service, and professional role, however, a lower number of professionals (27% vs 31%) with less than 15 years of experience reported to be aware of medical conditions that may be eligible for EI/ECSE services but not listed. Analysis based on professional role indicated that 24% of professionals whose role is to make referrals or provide medical diagnosis are aware of the medical conditions that may be eligible for services but not listed whereas 30% reported being unaware. Cronbach’s alpha score of 0.73 indicates higher internal consistencies among 4 items with binary response in the survey that were asked to explore the awareness level.

Qualitative analysis aligns with the survey response where respondents indicated that they learned about the diagnosed medical conditions that have a high probability of developmental delay, through their educational training or on-the-job. The majority of respondents (90%) said they have not seen the Oregon list of medical conditions before. EI professionals reported that programs have their own list of medical conditions. Lastly, healthcare professionals reported limited understanding of the purpose of the medical conditions list and a lack of understanding in who the list is for!

Research Question 2:

How are the categories of medical conditions that are listed in the Oregon State Plan policy guidance are understood and used by professionals in EI eligibility evaluations and assessments for birth to 5-year-old-children in Oregon? (IN PRACTICE)

Results were analyzed based on the overall response, profession, job status, years of service, and professional role. Consistent results were found across all groups indicating limited use of medical conditions list due to the lack of information. Overall analysis indicates that 77% of professionals ask for information on medical conditions from colleague to decide on EI referrals and eligibility evaluation, 58% have experienced a delay in EI eligibility evaluation due to lack of information, 57% have experienced a delay in timely receipt of EI services for children with diagnosed medical conditions, and 53% have looked for more information on medical conditions for EI eligibility determination. Sixty-eight percent reported that they have not skipped EI eligibility evaluation for children with diagnosed medical conditions that have a high probability of developmental delay.

Qualitative analysis provided similar insights where healthcare professionals reproted inconsistency in providing physician statement. They also reported limited understanding on how having a medical diagnosis and physician statement help the child, family, and the EI evaluation team. EI agencies have collated their own list of medical conditions in alignment with the state list, which may lead to county-to-county variation in the referral and evaluation process.

Research Question 3:

How and to what extent are EI eligibility evaluation and assessments for children with diagnosed medical conditions impacted during the covid-19 restrictions? (During COVID-19)

During the COVID-19 pandemic, 94% professionals reported to provide services remotely, 62% experienced delay in EI eligiblity determination for children with diagnosed medical conditions, and 48% have experienced decrease in number of children accessing healthcare and part C EI/ECSE services. A set of questions were asked only from EI/ECSE professionals to explore the state of uptake of EI/ECSE serivces during the pandemic where 77% reported to follow the 45-day timeline to do EI eligibility evaluation, assessments and IFSP development. Forty percent EI/ECSE professionals are using only tools that can be administered by parents during the COVID-19 pandemic and 30% reported to meet children and families in-person or in home visits.

Only 23% EI/ECSE professionals have reported to skip EI eligibility evaluation for children referred with written documentation of eligibility, and 9% reported to enroll children with interim IFSP where EI eligibility evaluation was not possible for children with diagnosed medical conditions during the COVID-19 pandemic. Low percentage of professionals reporting to skip EI eligibility evaluation and enrolling children with interim IFSP indicate lack of awareness or information among EI/ECSE professionals of federal regulations on children referred with diagnoed medical conditions that have a high probability of developmental delay or disability.

The qualitative analysis confirmed the above findings. As an urgent response to covid-19 pandemic, the EI/ECSE and healthcare professionals shifted to online platform for evaluations, assessments and other services. Specifically, professionals indicated that virtual platform is not suitable for delivering services to young children to carry out evaluations and assessments. In addition, virtual platform has put more burden on families to carry out observations with no technical expertise and lack of suitable materials. Some professionals also discussed that having a well-written medical conditions list and awareness among professionals would have benefitted some families who had to go through the eligibility determination process where they could have by-passed it during the covid-19 pandemic. Results also indicated that number of referrals were severely reduced due to the covid-19 pandemic.

Research Question 4:

What changes do EI professionals in Oregon recommend in the level of information provided about the categories of medical conditions that are likely to result in developmental delay? (VIEWS)

The analysis resulted in a consistent response from EI/ECSE and healthcare professionals in terms of changes they would prefer in medical conditions listed in the Oregon State Plan Policy guidance. Ninety percent reported having a more detailed and comprehensive list would help to decide when EI eligibility evaluations can be skipped and direct referrals are made for EI assessments and 86% percent agreed that a more detailed and comprehensive list of diagnosed medical conditions is needed in Oregon. Seventy-nine percent said a detailed and comprehensive list would have been helpful during the COVID-19 pandemic. Sixty-five percent said adequate information is not provided in the list of medical conditions that make children eligible for Part C EI/ECSE services in Oregon.

During the semi-structured interviews, we investigated views of participants on the current list and if they would like to suggest any changes. Respondents appreciated that there is a list of medical conditions in Oregon, however they greatly agreed in their replies that the current list is not seen by them and they do not refer it. The three most agreed recommendations are to have a more comprehensive, detailed list, widely circulated and presented in relevant forums on a regular basis and it should be embeded within the current system such as in the universal referral form and in the evaluation packet in EI programs to promote its appropriate utilization and increased understanding among professionals and families.

Conclusion

Row

Suggestion1

Problem: According to professionals people may think if condition is not on the list, it is not eligible becasue list contains limited information..

Statement:

“I suppose the trouble with A list is that there’s always those outlying diagnoses where, ’oh, it’s not on the list so it doesn’t fit. A list of specific conditions would be nice but with the caveat of it being like this isn’t necessarily an all-inclusive list.”- Speech and Language Pathologist

“I think more information would always be better. A lot of times we will reference this and if the condition is on there then you know it’s a lot easier for us to bring the child in but I mean there’s a lot of times where we have conditions that aren’t on this list and so then we kind of have to go and do our own research. So if we find research that points to the child having a condition or there’s been research studies that show that for the potential of developmental delay then we cite that kind of documentation and use that to bring them in if it’s not on this list.”- Physical Therapist

“One conditions was cleft lip and palate. So I actually don’t know if that’s on the list but if it’s not, certainly it should, because it has a huge developmental impact, even with mild cases.” – Speech and Language Pathologist

Recommendation:

Having more detailed, comprehensive list in simple language with more conditions that are almost 100% of the time eligible will be helpful. However, It would be helpful to have disclaimer that the list is not an all-inclusive list and that for more information a reliable web resource on medical conditions need to be provided.

Suggestion2

Problem: Lack of clarity in the purpose and utilization of the list? Is it for medical professionals, families or Early Interventional professionals?.

Statement:

“I think it’s very important for medical professionals to be sharing this with us, the teachers who should be looking out for these things and maybe even checking in with the parents and asking them if they have concerns.”- EI/ECSE professional.

“I would just send the referral and trust the EC cares would figure out if it’s inappropriate.”- Pediatrician

Recommendation:

Need more streamlined, clear presentation of information in simple langauge with instruction on how list can be utlized and implemented.

Suggestion3

Problem: Respondents varied greatly in their replies to making the list accessible for families.

Statement:

“I would love a list that I can share with a parent that they can understand without me needing to translate it alot. May be something in more layman’s terms.” (EI/ECSE professional)

“It’s certainly a very involved document and you need some sense of background to understand a lot of it but I think it’s important to be transparent with parents about this. I would think if any child might fall into any of these things a bit near, they should be provided to parents after birth or at the point which pediatricians or Health Providers are evaluating and assessing children and helping so this should be shared.” – (EI/ECSE professional)

“I don’t know how many parents would prefer this. I don’t know if it’s in super family-friendly language.” – EI/ECSE professional

“I don’t know if they had families look through it and given feedback I think that would be a good place to start.” -Physical Therapist

“It is always great to have information available for families that is appropriate and relevant and in family-friendly language. However, just the idea that the process of a physician statement is so different from a standard referral so most of the time when somebody referred for a physician or a med statement because of his medical condition. The doctor is saying here’s a program that will help you because we know things will go wrong. So it’s an active referral for developmental delay versus passive referral for a physician statement. So I don’t know if families would access it really all that much because they are already in the program by the time they would need it.” EI/ECSE specialist

Recommendation:

Parent groups should be consulted as part of the decision making process to get their feedback on how useful list would be for families.

Suggestion4

Problem: Current list is not accessible/visible/widely circulated. Only 10% interviewees had seen the Oregon list of medical conditions before the interview. Professionals were aware that children can be automatically eligible for EI services based on certain medical conditions, however, majority of participants (90%) have not seen the list before or referred in their professional role.

Statement:

“I’ve had a lot of presentations by EI folks and it’s never been presented to me so maybe when they go out to do presentations… they often go around to practices to talk about their services, maybe that should be part of what they talk about.” – Pediatrician

“ I don’t think any parents have ever seen anything like this. I don’t think parents really understand. A lot of parents don’t even understand what we are, what we do.” – EI/ECSE professional

“Having more people know about it. You know it’s surprising to me having worked in the field of early intervention for 22 years and I’ve never seen an actual list and if this is from the state of Oregon it seems like with all the paperwork that I do get and all the training that I do get it would be easy to be like…. Hey, here is a sheet of paper you should be aware of.”

Recommendation:

  1. List should be easily available electronically and widely distributed across pediatric centers and early intervention agencies in Oregon.

  2. An electronic list with a link for medical resource for more information should be provided.

  3. List can be published on popular websites such as Oregon Pediatric Society and Family Practice Society so that healthcare professionals do not miss this.

  4. List should be included in relevant presentations and forums regularly in Oregon.

Suggestion5

Problem: Current list is not embedded in the systems and practices in Oregon.

Statement:

“But they don’t list any on the referral form, they give examples like that they have autism but there is not the entire list there and it doesn’t refer to a list. So it could even just say look at this website, if you want to see the list of what we are asking for,“ – Pediatrician

“We have many children who are referred in, may be 18 months old or 2 or 3 or 3.5 years and we go ahead and then we start this standardized evaluation eligibility process. Never have I been cued to before I start to say bring this out and review something like this with the parent to make sure that the child doesn’t fit into one of these categories. We don’t ask the parent, especially during covid-19, we could by-pass the standardized evaluations.”- EI/ECSE specialist

Recommendation:

  1. List can be made available like a reference list on the universal referral form as an additional sheet. An online version can include pop up connected to the list of medical conditions would potentially open up the opportunity for medical professionals to understand and utilize it better.

  2. List can be included in the evaluation packet in EI/ECSE programs with an instruction to review these conditions before your proceed with a standardized evaluation.

  3. List should be included in orientation and induction of new recruits in healthcare and early intervention programs.

Column

Implementation Science

To have a long-term, consistent practice change for increased utilization, awareness and cross-sector uniformity (Part C Early Intervention and Healthcare) in the use of list of medical conditions for seamless identification, referral and uptake of Part C EI services for birth to 3-year-old children with diagnosed medical conditions, Oregon state need to adopt a focussed approach using the implementation science lens. Based on the in-depth data anlysis, we recommend use of Plan-DO-Study-Act (PDSA) process that encourages (i)planning the change(Plan), (ii) implementing the change on a small scale (Do), (iii) analyzing data and results(Study), and (iv) refining the changes based on the results(Act).

PLAN: To set up a working committee consisting of professionals from the EI evaluation team (physical therapist, speech and langauge therapist), pediatrician/care manager, a representative of parent group, and a state representative. The purpose of this committee is to devise a new comprehensive, detailed medical conditions list with appropriate information, presented in a user-friendly language and manner. Additionally, committee will also devise a focussed dissemination and implementation plan at the planning stage.

DO: Embed the new list in existing processes and practices including trainings, forums, orientation and induction in a selected EI program and a set of healthcare professionals on a small scale.

STUDY: Collect feedback and analyze the results.

ACT: Implement the recommended changes and go through the cycle again until intended results for sustainable use are achieved. Followed by state-wide implementation of the list of medical conditions in systems, processes, and practices.

Conclusion

Lack of awareness, access, and inconsistent utilization of the list of diagnosed medical conditions that give children direct access to Part C EI/ECSE services in the state of Oregon indicate that a more comprehensive and detailed list in a user-friendly language is needed. For a long-term sustainable use and seamless cross-sectoral referral and evaluation between Part C EI and healthcare professionals, the list need to be embedded in the existing systems and practices.

---
title: "Part C Eligibility Based on Medical Conditions in Oregon"
output: 
  flexdashboard::flex_dashboard:
    theme: sandstone
    orientation: columns
    vertical_layout: fill
    social: menu
    source_code: embed
---
<style>                     
.navbar {
  background-color:#003D79;
  border-color:white;
}
.navbar-brand {
color:white!important;
}

</style>   

<style type="text/css">

.chart-title {  /* chart_title  */
   font-size:12px

</style>



```{r setup, include=FALSE}
library(flexdashboard)
library(tidyverse)
library(rio)
library(here)
library(ggthemes)
library(ggpubr)
library(psych)
#install.packages("devtools")
library(devtools)
#remotes::install_github("wilkelab/cowplot")
#install.packages("colorspace", repos = "http://R-Forge.R-project.org")
#remotes::install_github("clauswilke/colorblindr")
library(colorspace)
library(ggplot2)
library(colorblindr)
library(scales)
#install.packages("showtext")
library(showtext)

library(gt)
library(kableExtra)
#install.packages("dygraphs")
library(dygraphs)
library(plotly)
#devtools::install_github("gadenbuie/xaringanthemer")
library(xaringanthemer)
library(kableExtra)

#install.packages("DiagrammeR")
library(DiagrammeR)
#install.packages('patchwork')
library(patchwork)


```

```{r, data}

data <- import(here::here("data", "ei_eligibility.xlsx"))


```



```{r xaringan-themer, include=FALSE, warning = FALSE}

style_mono_accent(
  base_color = "#DC322F", #  bright red
  inverse_background_color = "#002B36", # dark dark blue
  inverse_header_color = "#31b09e",     # light aqua green
  inverse_text_color = "#FFFFFF",       # white
  title_slide_background_color = "var(--base)",
  text_font_google = google_font("Kelly Slab","100", "100i"),
  header_font_google = google_font("Hind","100","100i"))

```

Descriptive
============================================

Column
-----------------------------------------------------------------------
### **About the study**
**Objective:**
The study aims to explore the awareness and use of the list of diagnosed medical conditions in Part C eligibility determination, from the perspective of the Early Intervention and Early Childhood Special Education (EI/ECSE), and healthcare professionals in the state of Oregon. 

**Methods:** 
The embedded mixed-method design was adopted where the qualitative (interviews) data were nested within the quantitative  (an electronic survey). Data from the survey and semi-structured interviews were collected simultaneously between November 2020 to January 2021. Responses were analyzed using descriptive statistics and thematic coding. 

**Results:** 
A total of 90 EI/ECSE and healthcare professionals responded to the electronic survey and 18 professionals attended the semi-structured interviews. The mixed-method analysis indicates that 74% of professionals are aware of the eligible medical conditions, however, when they were shown the Oregon list of medical conditions, they reported not to have seen the list before. The majority of professionals (77%) reported they seek help from colleagues to decide EI referrals and eligibility evaluation instead of using the list. 
As an urgent response to the covid-19 pandemic, professionals (94%) shifted to an online platform for evaluations, assessments, and other services. Professionals reported a severe decrease in the number of referrals (48%) and delay in EI eligibility determination (62%) for children with diagnosed medical conditions. 
Professionals greatly agreed in their replies that the current list does not contain adequate information (65%) and they do not refer to it for eligibility determination. The three most agreed recommendations are to have a more comprehensive, detailed list(86%) especially during   covid-19 (79%), widely circulated and embedded within the current system such as in the universal referral form and the evaluation packet in EI programs. 

**Conclusion:** 
Lack of awareness, access, and inconsistent use of the list of diagnosed medical conditions that give children direct access to Part C EI/ECSE services can lead to an imbalance in number and timely receipt of Part C EI services in Oregon. A comprehensive list in a user-friendly language is needed. The list needs to be embedded in the existing systems for its long-term sustainable use and seamless cross-sectoral referral and evaluation.

**Protocol: 09032020.003**

### **Descriptive analysis**

```{r}

# recoding the position to two categories: EI Professional and Healthcare Profession

data <- data %>%
  mutate(position = recode(position,
                           "Assistant teacher" = "EI/ECSE Personnel",
                           "EI coordinator" = "EI/ECSE Personnel",
                           "EI/ECSE Coordinator" = "EI/ECSE Personnel",
                           "EI/ECSE director" = "EI/ECSE Personnel",
                           "EI/ECSE specialist" = "EI/ECSE Personnel",
                           "EI/ECSE specialist autism" = "EI/ECSE Personnel",
                           "EI/ECSE specialist, & ECSE ABA Teacher" = "EI/ECSE Personnel",
                           "EI/ECSE specialist, EI coordinator" = "EI/ECSE Personnel",
                           "EI/ECSE specialist, EI/ECSE specialist autism" = "EI/ECSE Personnel",
                           "Eval Team Manager" = "EI/ECSE Personnel",
                           "Intake Specialist who works w/med statement referrals" = "EI/ECSE Personnel",
                           "Nurse practitioner" = "Healthcare Practitioner",
                           "Physical therapist, Evaluator-- evaluation team" = "EI/ECSE Personnel",
                           "Physician" = "Healthcare Practitioner",
                           "Physician assistant" = "Healthcare Practitioner",
                           "Program coordinator" = "EI/ECSE Personnel",
                           "Senior advisor" = "EI/ECSE Personnel",
                           "State EI/ECSE staff" = "EI/ECSE Personnel",
                           "EI professional" = "EI/ECSE Personnel",
                           "Healthcare professional" = "Healthcare Practitioner",
                           "Physical therapist" = "EI/ECSE Personnel",
                           "Occupational therapist" = "EI/ECSE Personnel",
                           "Speech & Language pathologist" = "EI/ECSE Personnel"))


# recoding the role to two categories:

data <- data %>%
  mutate(role = recode(role,
                       "Providing medical or developmental diagnosis." = "Making Referrals to EI and\n Providing Med. Diagnosis",
                       "Both, making referral for EI eligibility evaluation and providing medical diagnosis." = 
                         "Making Referrals to EI and\n Providing Med. Diagnosis",
                       "Making a referral for EI eligibility evaluation." = 
                         "Making Referrals to EI and\n Providing Med. Diagnosis",
                       "Carrying out EI eligibility evaluation." = "Part C EI/ECSE \nEligibility Evaluations"))

data <- data %>%
  mutate(years_service = recode(years_service,
                                "0-5" = "Less than 15 years",
                                "5-10" = "Less than 15 years",
                                "10-15" = "Less than 15 years",
                                "15 and above" = "Above 15 years"))

#data[sapply(data, is.character)] <- lapply(data[sapply(data,is.character)],as.factor)

table_pos <- data %>%
  select(position) %>%
  group_by(position) %>%
  count() %>%
  mutate(percent = round((n/90)*100, digits = 0))

table_job_status <- data %>%
  select(job_status) %>%
  group_by(job_status) %>%
  count() %>%
  mutate(percent = round((n/90)*100, digits = 0))

table_years_service <- data %>%
  select(years_service) %>%
  group_by(years_service) %>%
  count() %>%
  mutate(percent = round((n/90)*100, digits = 0))

table_role <- data%>%
  select(role) %>%
  group_by(role) %>%
  count() %>%
  mutate(percent = round((n/90)*100, digits = 0))


table_data <- bind_rows(table_pos, table_job_status, table_role, table_years_service) %>%
  rename("N" = "n",
         "Percent" = "percent")

table_data %>%
  unite(Participants, position, job_status, role, years_service, na.rm = TRUE) %>%
  kableExtra::kable() %>%
  kable_styling(full_width = F) %>%
  pack_rows("Profession", 1, 2, label_row_css = "background-color: #F0FFFF; color: black;") %>%
  pack_rows("Job Status", 3, 4, label_row_css = "background-color: #F0FFFF; color: black;") %>%
  pack_rows("Role", 5, 6, label_row_css = "background-color: #F0FFFF; color: black;") %>%
  pack_rows("Years of Service", 7, 8, label_row_css = "background-color:#F0FFFF; color: black;") %>%
  column_spec(1:3, color = "black")
  

```


Column {data-width=500}
-----------------------------------------------------------------------

### **Chart A: Survey participants across Oregon** 
Participants represented all 36 counties in Oregon. "Multiple" includes participants working in more than one county.

```{r}

county_data <- data %>%
  select(county) %>%
  mutate(county = recode(county,
                         "All" = "Multiple",
                         "Benton County, Lincoln County, Linn County" = "Multiple",
                         "Clackamas County, Clatsop County, Marion County, Multnomah County, Tilamook County, 
                         Washington County, Yamhill County" = "Multiple",
                         "Clackamas County, Multnomah County, Washington County" = "Multiple",
                         "Clackamas County, Multnomah County, Washington County, All" = "Multiple",
                         "Clatsop County, Columbia County, Tillamook County, Washington County" = "Multiple",
                         "Columbia County, Washington County" = "Multiple",
                         "Coos County, Curry County, Douglas County" = "Multiple",
                         "Crook County, Deschutes County, Jefferson County" = "Multiple",
                         "Deschutes County, Jefferson County" = "Multiple",
                         "Deschutes County, Klamath County" = "Multiple",
                         "Klamath County, Lake County" = "Multiple",
                         "Lane County, Linn County" = "Multiple",
                         "Marion County, Polk County" = "Multiple",
                         "Marion County, Polk County, Yamhill County" = "Multiple",
                         "Multnomah County, Washington County" = "Multiple",
                         "Clackamas County, Clatsop County, Marion County, Multnomah County, Tillamook County, Washington County, Yamhill County" = "Multiple")) %>%
  group_by(county) %>%
  count() %>%
  mutate(perc = round((n/90)*100, digits = 0))


county_data %>%
  ggplot(aes(x = perc, y = reorder(county, perc), fill = county)) +
  geom_vline(xintercept = c(10, 20, 30, 40), color = "grey75") +
  geom_col(fill = "gray75", alpha = 0.8) +
  geom_col(data = filter(county_data,
                         county == "Multiple"|
                         county == "Lane County"),
           fill = "#C55644") +
  #scale_x_continuous(labels = scales::percent) + 
  scale_x_continuous(breaks = pretty(county_data$perc), 
                    labels = paste0((pretty(county_data$perc)), "%"),
                    expand = c(0, 0)) +
  geom_text(aes(label = paste0(county_data$perc, "%")), color = "black", size = 4, hjust = -0.5) +
  theme_xaringan() +
  labs(x = "Percentage of respondents across Oregon",
       y = NULL,
       title = "Survey Participants Across Oregon") +
  theme(legend.position = "none",
        plot.title.position = "plot",
        plot.title = element_text(size = 15),
        axis.title.x = element_text(size = 15),
        axis.text.y = element_text(size = 12),
        axis.text.x = element_text(size = 12))
 
    
  

```




Awareness
=============================================

Column{data-height=50%, data-width=15%}
-----------------------------------------------------------------------

### Cronbach's alpha to measure how closely related set of questions are to indicate participants' awareness.

```{r}

valueBox(0.73,
         icon = "fas fa-thumbs-up")

```


### Survey participants fully completed this section.

```{r}

valueBox(90,
         icon = "ion-person-stalker")

```

Column{}
------------------------------------------------------------------

### Questions in the survey (Awareness)

```{r}

data%>%
  select(6:9) %>%
  pivot_longer(cols = starts_with("aware_"),
               names_to = "Questions",
               values_to = "Response") %>%
  group_by(Questions) %>%
  count() %>% 
  mutate("Questions" = recode(Questions, 
                              "aware_1_both" = "Q1: Are you aware of the medical conditions for children that have a high probability of significant developmental delay listed on Oregon State Plan Policy guidance?",
                              "aware_2_both" = "Q2: Are you aware that there are medical conditions for children that may qualify for EI eligibility evaluation and services that are not listed on Oregon State Plan Policy guidance?",
                              "aware_3_both" = "Q3: Are you aware that there are medical conditions for which minimum criteria need to be met to be eligible for EI services and a child’s disability does not need to be presently adversely affecting the child’s development?",
                              "aware_4_both" = "Q4: Are you aware of the medical documents and reports that contribute to the written documentation of eligibility for EI services based on the medical condition?"
                              )) %>%
  rename("N" = "n") %>%
  kableExtra::kable() %>%
  kable_styling(full_width = TRUE) %>%
  column_spec(1, width = "20cm", background = "#F0FFFF" )

```
### Analysis: Fisher's exact test

```{r}
aware_fisher <- data.frame(
  Ques = c("Q1:Aware of Med. Condition", 
           "Q2:Elig Med. Cond. Not Listed", 
           "Q3:Med. Cond. Presently Not Affecting",
           "Q4:Written Documentation of Eligiblity"),
  "EI/ECSE Personnel Vs Healthcare Practitioner" = c("0.0001***", "0.001***", "0.001***", "2.295"),
  "Full-time Vs Part-time" = c("0.01***", "0.588", "0.399", "0.548"),
  "Above 15 years of service Vs Less than 15 years" = c("0.226", "0.053", "0.651", "0.101"),
  "Part C EI/ECSE eligibility evaluations Vs Making referrals and providing diagnosis" = c("5.665", "0.055", "5.665", "6.305")
) 

aware_fisher %>%
  kableExtra::kable(col.names = c("Ques", "EI/ECSE Personnel\n Vs\nHealthcare Practitioner",
                                  "Full-time\nVs\nPart-time",
                                  "Above 15 years of service\n Vs \n Less than 15 years",
                                  "Part C EI/ECSE eligibility evaluations \n Vs \n Making referrals and providing diagnosis")) %>%
  kable_styling(full_width = FALSE) %>%
  footnote("***Fisher's exact test p<0.05 at 95% confidence interval")



```




Row {.tabset .tabset-fade}
-----------------------------------------------------------------------


### All participants




```{r}

aware_1 <- data %>%
  select(aware_1_both) %>%
  group_by(aware_1_both) %>%
  count() %>%
  rename("Aware of Med. Condition" = n)
 

aware_2 <- data%>%
  select(aware_2_both) %>%
  group_by(aware_2_both) %>%
  count() %>%
  rename("Elig Med. Cond. Not Listed" = n)
 

aware_3 <- data%>%
  select(aware_3_both) %>%
  group_by(aware_3_both) %>%
  count() %>%
  rename("Med. Cond. Presently \nNot Affecting" = n)
 

aware_4 <- data %>%
  select(aware_4_both) %>%
  group_by(aware_4_both) %>%
  count() %>%
  rename("Written Documentation \nof Eligibility" = n)
  
 

aware_plot <- bind_cols(aware_1, aware_2, aware_3, aware_4) %>%
  select(aware_1_both, "Aware of Med. Condition","Elig Med. Cond. Not Listed","Med. Cond. Presently \nNot Affecting",
         "Written Documentation \nof Eligibility") %>% 
  pivot_longer(cols = c("Aware of Med. Condition","Elig Med. Cond. Not Listed",
                        "Med. Cond. Presently \nNot Affecting", "Written Documentation \nof Eligibility"),
               names_to = "aware",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))


aware_plot <- aware_plot %>%
  rename("response" = "aware_1_both") %>%
  mutate(response = recode(response,
                           "No" = "Not aware",
                           "Yes" = "Aware"))

# For the divergign plot
aware_plot <- aware_plot %>%
mutate(percent = ifelse(response == "Aware",
                         percent, 
                       -1*percent))



#Plot AWareness: By Profession

ggplot(aware_plot,
       aes(x = percent, y = fct_reorder(aware, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(aware_plot$percent), 
                     labels = paste0(abs(pretty(aware_plot$percent)), "%")) +
  geom_text(aes(x = percent/2, y = aware, label = paste0(abs(aware_plot$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan() +
  ggtitle("Awareness: All Participants") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12, color = "black"),
         plot.title = element_text(color="red", size=12, hjust = 0.5),
         legend.text = element_text(size = 15)) +
  guides(fill = guide_legend(reverse = TRUE))

  

```



### By profession



```{r}

by_pos <- data %>%
  select(position, aware_1_both, aware_2_both, aware_3_both, aware_4_both)

pos_aware1 <- by_pos %>%
  select(position, aware_1_both) %>%
  group_by(position, aware_1_both) %>%
  count(aware_1_both) %>%
  rename("response" = "aware_1_both",
         "Aware of Med.\n Condition" = "n")


pos_aware2 <- by_pos %>%
  select(position, aware_2_both) %>%
  group_by(position, aware_2_both) %>%
  count(aware_2_both) %>%
  rename("response" = "aware_2_both",
         "Elig Med. Cond.\n Not Listed" = "n")
  

pos_aware3 <- by_pos %>%
  select(position, aware_3_both) %>%
  group_by(position, aware_3_both) %>%
  count(aware_3_both) %>%
  rename("response" = "aware_3_both",
         "Med. Cond. Presently\n Not Affecting" = "n")
 


pos_aware4 <- by_pos %>%
  select(position, aware_4_both) %>%
  group_by(position, aware_4_both) %>%
  count(aware_4_both) %>%
  rename("response" = "aware_4_both",
         "Written Documentation\n of Eligibility" = "n")



data_pos <- left_join(pos_aware1, pos_aware2, by = c("position", "response"))
data_pos <- left_join(data_pos, pos_aware3, by = c("position", "response"))
data_pos <- left_join(data_pos, pos_aware4, by = c("position", "response"))

data_pos <- data_pos %>%
  pivot_longer(cols = -c(position, response),
               names_to = "aware",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

data_pos <- data_pos %>%
  mutate(response = recode(response,
                           "Yes" = "Aware",
                           "No" = "Not aware")) %>%
  mutate(percent = ifelse(response == "Aware",
                          percent, 
                          -1*percent))


plot_aware_byprofession <- ggplot(data_pos,
       aes(x = percent, y = fct_reorder(aware, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.6) +
  facet_wrap(~position) +
  geom_vline(xintercept = 0, color = "#A0A0A0", linetype = "dashed") + #"red") +
  #colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(data_pos$percent),
                     labels = paste0(abs(pretty(data_pos$percent)), "%")) +
  geom_text(aes(x = percent/2, 
                y = aware, 
                label = paste0(abs(data_pos$percent), "%")), 
            color = "black", #"red", 
            size = 4) +
  #theme_xaringan() +
  theme_minimal() +
  
  ggtitle("(a) Awareness/knowledge: By Profession") +
  #labs(caption = "Figure 1") +
  theme(axis.title = element_blank(),
        panel.grid.minor = element_blank(),
        legend.title = element_blank(),
        legend.position = "bottom",
        legend.direction = "horizontal",
        plot.title.position = "plot",
        axis.text=element_text(size=12, color = "black"),
        plot.title = element_text(color= "black", #"red", 
                                  size=12, 
                                  hjust = 0.5),
        legend.text = element_text(size = 12),
        strip.background = element_blank(),
        strip.text = element_text(size = 12, #face="bold",
                                  color = "black", #"#d55e00", 
                                  hjust = 0.5)) +  #+
  guides(fill = guide_legend(reverse = TRUE)) +
  scale_fill_manual(values = c("#B8B8B8", "#D0D0D0"))
  

plot_aware_byprofession

```
```{r}
# save the plot in same directory.
png(file="plot_aware_byprofession.png")
plot_aware_byprofession
dev.off()
```

### By Job status

```{r}

status <- data%>%
  select(job_status, aware_1_both, aware_2_both, aware_3_both, aware_4_both)

status_aware1 <- status %>%
  select(job_status, aware_1_both) %>%
  group_by(job_status, aware_1_both) %>%
  count(aware_1_both) %>%
  rename("response" = "aware_1_both",
         "Aware of Med.\n Condition" = "n")


status_aware2 <- status %>%
  select(job_status, aware_2_both) %>%
  group_by(job_status, aware_2_both) %>%
  count(aware_2_both) %>%
  rename("response" = "aware_2_both",
         "Elig Med. Cond.\n Not Listed" = "n")
  

status_aware3 <- status %>%
  select(job_status, aware_3_both) %>%
  group_by(job_status, aware_3_both) %>%
  count(aware_3_both) %>%
  rename("response" = "aware_3_both",
         "Med. Cond. Presently\n Not Affecting" = "n")
 


status_aware4 <- status %>%
  select(job_status, aware_4_both) %>%
  group_by(job_status, aware_4_both) %>%
  count(aware_4_both) %>%
  rename("response" = "aware_4_both",
         "Written Documentation\n of Eligibility" = "n")



status_data <- left_join(status_aware1, status_aware2, by = c("job_status", "response"))
status_data <- left_join(status_data, status_aware3, by = c("job_status", "response"))
status_data <- left_join(status_data, status_aware4, by = c("job_status", "response"))

status_data <- status_data %>%
  pivot_longer(cols = -c(job_status, response),
               names_to = "aware",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))


# for the diverging plot

status_data <- status_data %>%
  mutate(response = recode(response,
                           "Yes" = "Aware",
                           "No" = "Not aware")) %>%
  mutate(percent = ifelse(response == "Aware",
                          percent, 
                         -1*percent))

ggplot(status_data,
       aes(x = percent, y = fct_reorder(aware, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~job_status) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(status_data$percent),
                     labels = paste0(abs(pretty(status_data$percent)), "%")) +
  geom_text(aes(x = percent/2, y = aware, label = paste0(abs(status_data$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan() +
  ggtitle("Awareness of Listed Medical Conditions for \nPart C EI/ECSE Services by Job Status") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12),
         plot.title = element_text(color="red", size=12, hjust = 0.5),
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(face="bold", size = 12, color = "#d55e00", hjust = 0.5)) + 
  guides(fill = guide_legend(reverse = TRUE))




```

### Years of service

```{r}

service_aware1 <- data %>%
  select(years_service, aware_1_both) %>%
  group_by(years_service, aware_1_both) %>%
  count(aware_1_both) %>%
  rename("response" = "aware_1_both",
         "Aware of Med.\n Condition" = "n")


service_aware2 <- data %>%
  select(years_service, aware_2_both) %>%
  group_by(years_service, aware_2_both) %>%
  count(aware_2_both) %>%
  rename("response" = "aware_2_both",
         "Elig Med. Cond.\n Not Listed" = "n")
  

service_aware3 <- data %>%
  select(years_service, aware_3_both) %>%
  group_by(years_service, aware_3_both) %>%
  count(aware_3_both) %>%
  rename("response" = "aware_3_both",
         "Med. Cond. Presently\n Not Affecting" = "n")
 
service_aware4 <- data %>%
  select(years_service, aware_4_both) %>%
  group_by(years_service, aware_4_both) %>%
  count(aware_4_both) %>%
  rename("response" = "aware_4_both",
         "Written Documentation\n of Eligibility" = "n")

service_data <- left_join(service_aware1, service_aware2, by = c("years_service", "response"))
service_data <- left_join(service_data, service_aware3, by = c("years_service", "response"))
service_data <- left_join(service_data, service_aware4, by = c("years_service", "response"))


service_data <- service_data %>%
  pivot_longer(cols = -c(years_service, response),
               names_to = "aware",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

service_data <- service_data %>%
  mutate(response = recode(response,
                           "Yes" = "Aware",
                           "No" = "Not aware")) %>%
  mutate(percent = ifelse(response == "Aware",
                          percent, 
                         -1*percent))

ggplot(service_data,
       aes(x = percent, y = fct_reorder(aware, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~years_service) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(service_data$percent),
                     labels = paste0(abs(pretty(service_data$percent)), "%")) +
  geom_text(aes(x = percent/2, y = aware, label = paste0(abs(service_data$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan() +
  ggtitle("Awareness of Listed Medical Conditions for \nPart C EI/ECSE Services by Job Status") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12,),
         plot.title = element_text(color="red", size=12, hjust = 0.5),
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(face="bold", size = 12, color = "#d55e00", hjust = 0.5)) +
  guides(fill = guide_legend(reverse = TRUE)) 
  

```

### By professional role

```{r}

role_data <- data%>%
  select(role, aware_1_both, aware_2_both, aware_3_both, aware_4_both)

role_data <- role_data %>%
  mutate(role = recode(role,
                       "Providing medical or developmental diagnosis." = "Making Referrals to EI and\n Providing Med. Diagnosis",
                       "Both, making referral for EI eligibility evaluation and providing medical diagnosis." = 
                         "Making Referrals to EI and\n Providing Med. Diagnosis",
                       "Making a referral for EI eligibility evaluation." = 
                         "Making Referrals to EI and\n Providing Med. Diagnosis",
                       "Carrying out EI eligibility evaluation." = "Part C EI/ECSE \nEligibility Evaluations"))

role_aware1 <- role_data %>%
  select(role, aware_1_both) %>%
  group_by(role, aware_1_both) %>%
  count(aware_1_both) %>%
  rename("response" = "aware_1_both",
         "Aware of Med.\nCondition" = "n")

role_aware2 <- role_data %>%
  select(role, aware_2_both) %>%
  group_by(role, aware_2_both) %>%
  count(aware_2_both) %>%
  rename("response" = "aware_2_both",
         "Elig Med. Cond.\n Not Listed" = "n")
  

role_aware3 <- role_data %>%
  select(role, aware_3_both) %>%
  group_by(role, aware_3_both) %>%
  count(aware_3_both) %>%
  rename("response" = "aware_3_both",
         "Med. Cond. Presently\n Not Affecting" = "n")
 


role_aware4 <- role_data %>%
  select(role, aware_4_both) %>%
  group_by(role, aware_4_both) %>%
  count(aware_4_both) %>%
  rename("response" = "aware_4_both",
         "Written Documentation\n of Eligibility" = "n")


role_data <- left_join(role_aware1, role_aware2, by = c("role", "response"))
role_data <- left_join(role_data, role_aware3, by = c("role", "response"))
role_data <- left_join(role_data, role_aware4, by = c("role", "response"))

role_data <- role_data %>%
  pivot_longer(cols = -c(role, response),
               names_to = "aware",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

role_data <- role_data %>%
  mutate(response = recode(response,
                           "Yes" = "Aware",
                           "No" = "Not aware")) %>%
  mutate(percent = ifelse(response == "Aware",
                          percent, 
                         -1*percent))

ggplot(role_data,
       aes(x = percent, y = fct_reorder(aware, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~role) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(role_data$percent),
                     labels = paste0(abs(pretty(role_data$percent)), "%")) +
  geom_text(aes(x = percent/2, y = aware, label = paste0(abs(role_data$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan() +
  ggtitle("Awareness of Listed Medical Conditions for Part C EI/ECSE \nServices by Professional Role") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12),
         plot.title = element_text(color="red", size=12, hjust = 0.5),
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(size = 12, color = "#d55e00", hjust = 0.5)) + 
  guides(fill = guide_legend(reverse = TRUE))


```




In-practice
============================================


Column{data-height=50%, data-width=15%}
-----------------------------------------------------------------------


### Survey participants fully completed this section.

```{r}

valueBox(90,
         icon = "ion-person-stalker")

```

Column{}
------------------------------------------------------------------

### Questions in the survey (In-Practice)

```{r}

data%>%
  select(10:14) %>%
  pivot_longer(cols = ends_with("_both"),
               names_to = "Questions",
               values_to = "Response") %>%
  group_by(Questions) %>%
  count() %>% 
  mutate("Questions" = recode(Questions, 
                              "skipped_elig_both" = "Q: Have you ever referred a child directly for IFSP development based on a diagnosed medical condition, without doing the preliminary step of eligibility evaluation?",
                              "delay_elig_both" = "Q: Have you ever experienced a delay in the EI eligibility determination process due to a lack of information or clarity on a child’s medical condition?",
                              "difficulty_access_both" = "Q: Have you come across children with a diagnosed medical condition who could not access EI eligibility evaluation and other services on time?",
                              "help_colleague_both" = "Have you ever asked another professional for information on medical conditions to help evaluate a child’s EI eligibility?",
                              "searched_medcond_both" = "Q: Have you ever searched for information on medical conditions to make EI eligibility determination?")) %>%
  rename("N" = "n") %>%
  kableExtra::kable() %>%
  kable_styling(full_width = TRUE) %>%
  column_spec(1, width = "20cm", background = "#F0FFFF" )

```





Row {.tabset .tabset-fade}
-----------------------------------------------------------------------

### By all participants

```{r}




prac_1_all <- data %>%
  select(skipped_elig_both) %>%
  group_by(skipped_elig_both) %>%
  count() %>%
  rename("Skipped EI Eligibility Eval" = n,
         "response" = "skipped_elig_both" )
 

prac_2_all <- data%>%
  select(delay_elig_both) %>%
  group_by(delay_elig_both) %>%
  count() %>%
  rename("Delay in EI Eligibility Eval" = n,
         "response" = "delay_elig_both")
 

prac_3_all <- data%>%
  select(difficulty_access_both) %>%
  group_by(difficulty_access_both) %>%
  count() %>%
  rename("Difficulty access services" = n,
         "response" = "difficulty_access_both")
 

prac_4_all <- data %>%
  select(help_colleague_both) %>%
  group_by(help_colleague_both) %>%
  count() %>%
  rename("Ask for help from colleague" = n,
         "response" = "help_colleague_both")
  
 
prac_5_all <- data %>%
  select(searched_medcond_both) %>%
  group_by(searched_medcond_both) %>%
  count() %>%
  rename("Searched for eligible medical \n conditions" = n,
         "response" = "searched_medcond_both")


practice_data <- left_join(prac_1_all, prac_2_all, by = "response")
practice_data <- left_join(practice_data, prac_3_all, by =  "response")
practice_data <- left_join(practice_data, prac_4_all, by = "response")
practice_data <- left_join(practice_data, prac_5_all, by="response")

practice_data <- practice_data %>%
  pivot_longer(cols = -c("response"),
               names_to = "practice",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

practice_data <- practice_data %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))


#Plot AWareness: By Profession

In_practice_all<-ggplot(practice_data,
       aes(x = percent, y = fct_reorder(practice, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.6) +
  geom_vline(xintercept = 0, color = "#A0A0A0", linetype = "dashed") + #"red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(practice_data$percent), 
                     labels = paste0(abs(pretty(practice_data$percent)), "%")) +
  geom_text(aes(x = percent/2, y = practice, label = paste0(abs(practice_data$percent), "%")), 
            color = "black", #"red", 
            size = 4) +
 # theme_xaringan() +
  theme_minimal() +
  ggtitle("(b) In Practice: All Participants") +
 # labs(caption = "Figure 2") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12, color = "black"),
         plot.title = element_text(color= "black",size=12, hjust = 0.5), # "red", size=12, hjust = 0.5),
         legend.text = element_text(size = 12)) +
  guides(fill = guide_legend(reverse = FALSE)) +
  scale_fill_manual(values = c("#D0D0D0", "#B8B8B8"))

  In_practice_all
  
```

```{r}
# save the plot in same directory.
pdf(file="InPractice_All.pdf")
In_practice_all
dev.off()
```

### By profession

```{r, fig.width = 8}


prac_1_pro <- data %>%
  select(position, skipped_elig_both) %>%
  group_by(position, skipped_elig_both) %>%
  count() %>%
  rename("Skipped EI Eligibility Eval" = n,
         "response" = "skipped_elig_both" )
 

prac_2_pro <- data%>%
  select(position, delay_elig_both) %>%
  group_by(position, delay_elig_both) %>%
  count() %>%
  rename("Delay in EI Eligibility Eval" = n,
         "response" = "delay_elig_both")
 

prac_3_pro <- data%>%
  select(position, difficulty_access_both) %>%
  group_by(position, difficulty_access_both) %>%
  count() %>%
  rename("Difficulty access services" = n,
         "response" = "difficulty_access_both")
 

prac_4_pro <- data %>%
  select(position, help_colleague_both) %>%
  group_by(position, help_colleague_both) %>%
  count() %>%
  rename("Ask for help from colleague" = n,
         "response" = "help_colleague_both")
  
 
prac_5_pro <- data %>%
  select(position, searched_medcond_both) %>%
  group_by(position, searched_medcond_both) %>%
  count() %>%
  rename("Searched for eligible\n medical conditions" = n,
         "response" = "searched_medcond_both")


practice_data_pro <- left_join(prac_1_pro, prac_2_pro, by = c("position", "response"))
practice_data_pro <- left_join(practice_data_pro, prac_3_pro, by =  c("position", "response"))
practice_data_pro <- left_join(practice_data_pro, prac_4_pro, by = c("position", "response"))
practice_data_pro <- left_join(practice_data_pro, prac_5_pro, by = c("position", "response"))


practice_data_pro <- practice_data_pro %>%
  pivot_longer(cols = -c("position", "response"),
               names_to = "practice",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

practice_data_pro <- practice_data_pro %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))


#Plot AWareness: By Profession

In_practice_by_profession <- ggplot(practice_data_pro,
       aes(x = percent, y = fct_reorder(practice, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~position) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") + 
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(practice_data_pro$percent), 
                     labels = paste0(abs(pretty(practice_data_pro$percent)), "%")) +
  geom_text(aes(x = percent/2, y = practice, label = paste0(abs(practice_data_pro$percent), "%")), color = "black", #"red", 
            size = 3) +
  theme_xaringan() +
  ggtitle("In Practice (By Profession)") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12, color = "black"),
         plot.title = element_text(color= "black", size=15, hjust = 0.5), #"red"
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(face="bold", size = 15, color = "#d55e00", hjust = 0.5)) + #d55e00", hjust = 0.5)) +
  guides(fill = guide_legend(reverse = FALSE))
 


In_practice_by_profession

```

```{r}
# save the plot in same directory.
pdf(file="Inpractice_byprofession.pdf")
In_practice_by_profession
dev.off()
```

### By job status

```{r, fig.width=8}

prac_1_job <- data %>%
  select(job_status, skipped_elig_both) %>%
  group_by(job_status, skipped_elig_both) %>%
  count() %>%
  rename("Skipped EI Eligibility Eval" = n,
         "response" = "skipped_elig_both" )
 

prac_2_job <- data%>%
  select(job_status, delay_elig_both) %>%
  group_by(job_status, delay_elig_both) %>%
  count() %>%
  rename("Delay in EI Eligibility Eval" = n,
         "response" = "delay_elig_both")
 

prac_3_job <- data%>%
  select(job_status, difficulty_access_both) %>%
  group_by(job_status, difficulty_access_both) %>%
  count() %>%
  rename("Difficulty access services" = n,
         "response" = "difficulty_access_both")
 

prac_4_job <- data %>%
  select(job_status, help_colleague_both) %>%
  group_by(job_status, help_colleague_both) %>%
  count() %>%
  rename("Ask for help from colleague" = n,
         "response" = "help_colleague_both")
  
 
prac_5_job <- data %>%
  select(job_status, searched_medcond_both) %>%
  group_by(job_status, searched_medcond_both) %>%
  count() %>%
  rename("Searched for eligible\n medical conditions" = n,
         "response" = "searched_medcond_both")


practice_data_job <- left_join(prac_1_job, prac_2_job, by = c("job_status", "response"))
practice_data_job <- left_join(practice_data_job, prac_3_job, by =  c("job_status", "response"))
practice_data_job <- left_join(practice_data_job, prac_4_job, by = c("job_status", "response"))
practice_data_job <- left_join(practice_data_job, prac_5_job, by = c("job_status", "response"))


practice_data_job <- practice_data_job %>%
  pivot_longer(cols = -c("job_status", "response"),
               names_to = "practice",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

practice_data_job <- practice_data_job %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))


#Plot AWareness: By Profession

ggplot(practice_data_job,
       aes(x = percent, y = fct_reorder(practice, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~job_status) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(practice_data_job$percent), 
                     labels = paste0(abs(pretty(practice_data_job$percent)), "%")) +
  geom_text(aes(x = percent/2, y = practice, label = paste0(abs(practice_data_job$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan() +
  ggtitle("Use of Listed Medical Conditions By Job Status") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12),
         plot.title = element_text(color="red", size=15, hjust = 0.5),
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(face="bold", size = 15, color = "#d55e00", hjust = 0.5)) +
  guides(fill = guide_legend(reverse = FALSE))



```



### By years of service

```{r, fig.width=8}

prac_1_service <- data %>%
  select(years_service, skipped_elig_both) %>%
  group_by(years_service, skipped_elig_both) %>%
  count() %>%
  rename("Skipped EI Eligibility Eval" = n,
         "response" = "skipped_elig_both" )
 

prac_2_service <- data%>%
  select(years_service, delay_elig_both) %>%
  group_by(years_service, delay_elig_both) %>%
  count() %>%
  rename("Delay in EI Eligibility Eval" = n,
         "response" = "delay_elig_both")
 

prac_3_service <- data%>%
  select(years_service, difficulty_access_both) %>%
  group_by(years_service, difficulty_access_both) %>%
  count() %>%
  rename("Difficulty access services" = n,
         "response" = "difficulty_access_both")
 

prac_4_service <- data %>%
  select(years_service, help_colleague_both) %>%
  group_by(years_service, help_colleague_both) %>%
  count() %>%
  rename("Ask for help from colleague" = n,
         "response" = "help_colleague_both")
  
 
prac_5_service <- data %>%
  select(years_service, searched_medcond_both) %>%
  group_by(years_service, searched_medcond_both) %>%
  count() %>%
  rename("Searched for eligible\n medical conditions" = n,
         "response" = "searched_medcond_both")


practice_data_service <- left_join(prac_1_service, prac_2_service, by = c("years_service", "response"))
practice_data_service <- left_join(practice_data_service, prac_3_service, by =  c("years_service", "response"))
practice_data_service <- left_join(practice_data_service, prac_4_service, by = c("years_service", "response"))
practice_data_service <- left_join(practice_data_service, prac_5_service, by = c("years_service", "response"))


practice_data_service <- practice_data_service %>%
  pivot_longer(cols = -c("years_service", "response"),
               names_to = "practice",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

practice_data_service <- practice_data_service %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))


#Plot AWareness: By Profession

ggplot(practice_data_service,
       aes(x = percent, y = fct_reorder(practice, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~years_service) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(practice_data_job$percent), 
                     labels = paste0(abs(pretty(practice_data_job$percent)), "%")) +
  geom_text(aes(x = percent/2, y = practice, label = paste0(abs(practice_data_job$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan() +
  ggtitle("Use of Listed Medical Conditions By Years of Service") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12),
         plot.title = element_text(color="red", size=15, hjust = 0.5),
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(face="bold", size = 15, color = "#d55e00", hjust = 0.5)) +
  guides(fill = guide_legend(reverse = FALSE))


```

### By professional role

```{r, fig.width=8}


prac_1_role <- data %>%
  select(role, skipped_elig_both) %>%
  group_by(role, skipped_elig_both) %>%
  count() %>%
  rename("Skipped EI Eligibility Eval" = n,
         "response" = "skipped_elig_both" )
 

prac_2_role <- data%>%
  select(role, delay_elig_both) %>%
  group_by(role, delay_elig_both) %>%
  count() %>%
  rename("Delay in EI Eligibility Eval" = n,
         "response" = "delay_elig_both")
 

prac_3_role <- data%>%
  select(role, difficulty_access_both) %>%
  group_by(role, difficulty_access_both) %>%
  count() %>%
  rename("Difficulty access services" = n,
         "response" = "difficulty_access_both")
 

prac_4_role <- data %>%
  select(role, help_colleague_both) %>%
  group_by(role, help_colleague_both) %>%
  count() %>%
  rename("Ask for help from colleague" = n,
         "response" = "help_colleague_both")
  
 
prac_5_role <- data %>%
  select(role, searched_medcond_both) %>%
  group_by(role, searched_medcond_both) %>%
  count() %>%
  rename("Searched for eligible\n medical conditions" = n,
         "response" = "searched_medcond_both")


practice_data_role <- left_join(prac_1_role, prac_2_role, by = c("role", "response"))
practice_data_role <- left_join(practice_data_role, prac_3_role, by =  c("role", "response"))
practice_data_role <- left_join(practice_data_role, prac_4_role, by = c("role", "response"))
practice_data_role <- left_join(practice_data_role, prac_5_role, by = c("role", "response"))


practice_data_role <- practice_data_role %>%
  pivot_longer(cols = -c("role", "response"),
               names_to = "practice",
               values_to = "count") %>%
  mutate(percent = round((count/90)*100, digits = 0))

# for the diverging plot

practice_data_role <- practice_data_role %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))


#Plot AWareness: By Profession

ggplot(practice_data_role,
       aes(x = percent, y = fct_reorder(practice, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~role) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(practice_data_job$percent), 
                     labels = paste0(abs(pretty(practice_data_job$percent)), "%")) +
  geom_text(aes(x = percent/2, y = practice, label = paste0(abs(practice_data_job$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan_inverse() +
  ggtitle("Use of Listed Medical Conditions By Professional Role") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12),
         plot.title = element_text(color="red", size=15, hjust = 0.5),
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(face="bold", size = 12, color = "#d55e00", hjust = 0.5)) +
  guides(fill = guide_legend(reverse = FALSE))


```


During Covid-19
==========================================

Column{data-height=50%, data-width=15%}
-----------------------------------------------------------------------

### Total number of respondents fully completed this section of the survey.

```{r}

valueBox(87,
         icon = "ion-person-stalker")

```


### total number of EI/ECSE professionals fully completed this section of the survey.

```{r}

valueBox(37,
          icon = "ion-person-stalker")

```

Column {}
-----------------------------------------------------------------------

### Questions in the survey(COVID-19: For all participants)

```{r, questions_covid19both}


data%>%
  select(elig_remote_both, delay_covid_both, decrease_both) %>%
  drop_na() %>%
  pivot_longer(cols = ends_with("_both"),
               names_to = "Questions",
               values_to = "Response") %>%
  group_by(Questions) %>%
  count() %>% 
  mutate("Questions" = recode(Questions,
                              "elig_remote_both" = "Q: Are you conducting EI eligibility evaluations remotely through phone and videoconferencing for children during Covid-19?",
                              "delay_covid_both" = "Q: Have you experienced a delay in EI eligibility evaluation and assessment for any child diagnosed with a medical condition during Covid-19?",
                              "decrease_both" = "Q: Have you experienced a decrease in the number of children and families in need of referral for EI eligibility evaluation or medical diagnosis during Covid-19?")) %>%
  rename("N" = "n") %>%
  kableExtra::kable() %>%
  kable_styling(full_width = TRUE) %>%
  column_spec(1, width = "20cm", background = "#F0FFFF" )


```

### Questions in the survey(COVID-19: For EI/ECSE participants only)

```{r, questions_covid_ei}

data%>%
  select(`45_day`, tools_ei, hv_ei, skip_covid_ei, enrolled_ifsp_ei) %>%
  drop_na() %>%
  pivot_longer(cols = c(`45_day`, tools_ei, hv_ei, skip_covid_ei, enrolled_ifsp_ei),
               names_to = "Questions",
               values_to = "Response") %>%
  group_by(Questions) %>%
  count() %>% 
  mutate("Questions" = recode(Questions,
                              "45_day" = "Q: Are you following the 45-day timeline to do EI eligibility evaluation, assessments, and IFSP development for children during Covid-19?",
                              "tools_ei" = "Q: Are you using only tools that can be administered by parents to do EI evaluations for children during Covid-19?",
                              "hv_ei" = "Q: Have you met children or families in-person or made home visits for EI eligibility during Covid-19?",
                              "skip_covid_ei" = "Q: Have you skipped EI eligibility evaluation for children with a diagnosed medical condition who have written documentation of eligibility from their physician during Covid-19?",
                              "enrolled_ifsp_ei" = "Q: Have you enrolled children with a diagnosed medical condition with interim IFSP where EI eligibility evaluation was not possible during Covid-19?")) %>%
  rename("N" = "n") %>%
  kableExtra::kable() %>%
  kable_styling(full_width = TRUE) %>%
  column_spec(1, width = "20cm", background = "#F0FFFF" )



```

Row {.tabset .tabset-fade}
-----------------------------------------------------------------------

### By all participants

```{r, covid_both}

data_covid <- data %>%
  select(position, elig_remote_both, decrease_both, delay_covid_both, 
         challenges_covid_med, `45_day`, tools_ei, hv_ei, skip_covid_ei, enrolled_ifsp_ei)

cov_med_ei <- data_covid %>%
  select(elig_remote_both, decrease_both, delay_covid_both) %>%
  drop_na()

cov_med_ei_1 <- cov_med_ei %>%
  select(elig_remote_both) %>%
  group_by(elig_remote_both) %>%
  count(elig_remote_both) %>%
  rename("response" = "elig_remote_both",
         "Remote service delivery" = "n")

cov_med_ei_2 <- cov_med_ei %>%
  select(decrease_both) %>%
  group_by(decrease_both) %>%
  count(decrease_both) %>%
  rename("response" = "decrease_both",
         "Decrease in numbers" = "n")
  
cov_med_ei_3 <- cov_med_ei %>%
  select(delay_covid_both) %>%
  group_by(delay_covid_both) %>%
  count(delay_covid_both) %>%
  rename("response" = "delay_covid_both",
         "Delay in EI\n eligibility evaluation" = "n")

cov_med_ei <- left_join(cov_med_ei_1, cov_med_ei_2, by = c("response"))
cov_med_ei <- left_join(cov_med_ei, cov_med_ei_3, by = c("response"))

cov_med_ei <- cov_med_ei %>%
  pivot_longer(cols = -c(response),
               names_to = "during_pandemic",
               values_to = "count") %>%
  mutate(percent = round((count/85)*100, digits = 0))
  
# for the diverging plot

cov_med_ei <- cov_med_ei %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))

covid_allparticipants <- ggplot(cov_med_ei,
       aes(x = percent, y = fct_reorder(during_pandemic, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(cov_med_ei$percent),
                     labels = paste0(abs(pretty(cov_med_ei$percent)), "%")) +
  geom_text(aes(x = percent/2, y = during_pandemic, label = paste0(abs(cov_med_ei$percent), "%")), color = "red", 
            size = 4) +
  #facet_wrap(~position) +
  #theme_xaringan_inverse() +
  theme_xaringan() +
  ggtitle("During COVID-19 Pandemic (All Participants)") +
#  labs(caption = "Figure 3") +
  theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12, color = "black"),
         plot.title = element_text(color="red", size=15, hjust = 0.5),
         legend.text = element_text(size = 15))
        
covid_allparticipants

```

```{r}
# save the plot in same directory.
pdf(file="covid_allparticipants.pdf")
covid_allparticipants
dev.off()
```



### By EI/ECSE Professionals

```{r, covid_ei}

data_onepro <- data_covid %>%
  select(position,`45_day`, tools_ei, hv_ei, skip_covid_ei, enrolled_ifsp_ei) %>%
  filter(position == "EI/ECSE Personnel") %>%
  drop_na()

tools <- data_onepro %>%
  select(tools_ei) %>%
  drop_na() %>% #47
  group_by(tools_ei) %>%
  count() %>%
  rename("response" = "tools_ei",
         "Using parent-administered\n tool only" = "n")
  
Met_in_person <- data_onepro %>%
  select(hv_ei) %>%
  drop_na() %>% #47
  group_by(hv_ei) %>%
  count() %>%
  rename("response" = "hv_ei",
         "Met children in person or made\n homevisits" = "n")
  

timeline <- data_onepro %>%
  select(`45_day`) %>%
  drop_na() %>%        #41
  group_by(`45_day`) %>%
  count() %>%
  rename("response" = "45_day",
         "Following 45-day timeline" = "n")

skip <- data_onepro %>%
  select(skip_covid_ei) %>%
  drop_na() %>%         #38
  group_by(skip_covid_ei) %>%
  count() %>%
  rename("response" = skip_covid_ei,
         "Skipped evaluation for children\n with written documentation" = "n")
  
enrolled <- data_onepro %>%
  select(enrolled_ifsp_ei) %>%
  drop_na() %>%  #38
  group_by(enrolled_ifsp_ei) %>%
  count() %>%
  rename("response" = "enrolled_ifsp_ei",
         "Enrolled children with interim \nIFSP" = "n")

covid_only_ei <- left_join(tools, Met_in_person, by = c("response"))
covid_only_ei <- left_join(covid_only_ei, timeline, by = c("response"))
covid_only_ei <- left_join(covid_only_ei, skip, by = c("response"))
covid_only_ei <- left_join(covid_only_ei, enrolled, by = c("response"))


covid_only_ei <- covid_only_ei %>%
  pivot_longer(cols = c("Using parent-administered\n tool only",
                        "Met children in person or made\n homevisits",
                        "Following 45-day timeline",
                        "Skipped evaluation for children\n with written documentation",
                        "Enrolled children with interim \nIFSP"),
               names_to = "during_pandemic",
               values_to = "count") %>%
  mutate(percent = round((count/37)*100, digits = 0))



# for the diverging plot

covid_only_ei <- covid_only_ei %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))


covid_eiecse <- ggplot(covid_only_ei,
       aes(x = percent, y = fct_reorder(during_pandemic, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.6) +
  geom_vline(xintercept = 0, color = "#A0A0A0", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(covid_only_ei$percent),
                     labels = paste0(abs(pretty(covid_only_ei$percent)), "%")) +
  geom_text(aes(x = percent/2, y = during_pandemic, label = paste0(abs(covid_only_ei$percent),
                                                                   "%")), color = "black", 
            size = 4) +
  
  #facet_wrap(~position) +
# theme_xaringan_inverse() +
  #theme_xaringan() + 
  theme_minimal() +
  ggtitle("(c) Eligibility Evaluation During COVID-19: Part C EI/ECSE Personnel") +
  #labs(caption = "Figure 4") +
  theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12, color = "black"),
         plot.title = element_text(color="black", size=12, hjust = 0.5),
         legend.text = element_text(size = 12)) +
  scale_fill_manual(values = c("#D0D0D0", "#B8B8B8"))
        
covid_eiecse

```
```{r}
 # save the plot in same directory.
pdf(file="covid_eiecse.pdf")
covid_eiecse
dev.off()

```


Views
=========================================

Column{data-height=50%, data-width=20%}
-----------------------------------------------------------------------

### Cronbach's alpha to measure how closely related set of questions are to indicate participants' views on changes they expect in listed medical conditions.

```{r}

valueBox(0.64,
         icon = "fas fa-thumbs-up")

```


### Total number of respondents fully completed this section of the survey.

```{r}

valueBox(81,
         icon = "ion-person-stalker")

```


Column {data-width=300}
-----------------------------------------------------------------------


```{r}

### Questions

data%>%
  select(15:18) %>%
  drop_na() %>%
  pivot_longer(cols = ends_with("_both"),
               names_to = "Questions",
               values_to = "Response") %>%
  group_by(Questions) %>%
  count() %>% 
  mutate("Questions" = recode(Questions,
                              "detail_help_both" = "Q: In your opinion, would a more detailed and comprehensive list of medical conditions help to decide when EI eligibility evaluations can be skipped and direct referral be made for EI assessment?",
                              "detail_need_both" = "Q: In your opinion, is having a more detailed and comprehensive medical conditions list for EI eligibility determination in Oregon needed?",
                              "detail_covid_both" = "Q: In your opinion, a more detailed and comprehensive medical conditions list would have been helpful during Covid-19?",
                              "info_adequate_both" = "Q: Do you think that information on medical conditions listed in the Oregon State Plan policy guidance is adequate for deciding to carry out an EI eligibility evaluation?")) %>%
  rename("N" = "n") %>%
  kableExtra::kable() %>%
  kable_styling(full_width = TRUE) %>%
  column_spec(1, width = "15cm", background = "#F0FFFF" )
  
```



Row {.tabset .tabset-fade}
-----------------------------------------------------------------------



### By All Participants

```{r, change_all}

data_rec_overall <- data %>%
  select(info_adequate_both, detail_need_both, detail_help_both, detail_covid_both) %>%
  drop_na()


rec1 <- data_rec_overall %>%
  select(info_adequate_both) %>%
  group_by(info_adequate_both) %>%
  count(info_adequate_both) %>%
  rename("response" = "info_adequate_both",
         "Adequate information\n is given" = "n")


rec2 <- data_rec_overall %>%
  select(detail_need_both) %>%
  group_by(detail_need_both) %>%
  count(detail_need_both) %>%
  rename("response" = "detail_need_both",
         "Detailed list\nis needed" = "n")
  

rec3 <- data_rec_overall %>%
  select(detail_help_both) %>%
  group_by(detail_help_both) %>%
  count(detail_help_both) %>%
  rename("response" = "detail_help_both",
         "Detailed list\n would be helpful" = "n")
 


rec4 <- data_rec_overall %>%
  select(detail_covid_both) %>%
  group_by(detail_covid_both) %>%
  count(detail_covid_both) %>%
  rename("response" = "detail_covid_both",
         "Detailed list during\n Covid-19 is needed" = "n")



rec_overall <- left_join(rec1, rec2, by = "response")
rec_overall <- left_join(rec_overall, rec3, by = "response")
rec_overall <- left_join(rec_overall, rec4, by = "response")


rec_overall <- rec_overall %>%
  pivot_longer(cols = -response,
               names_to = "information",
               values_to = "count") %>%
  mutate(percent = round((count/81)*100, digits = 0))


# for the diverging plot

rec_overall <- rec_overall %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))

ggplot(rec_overall,
       aes(x = percent, y = fct_reorder(information, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(rec_overall$percent),
                     labels = paste0(abs(pretty(rec_overall$percent)), "%")) +
  geom_text(aes(x = percent/2, y = information, label = paste0(abs(rec_overall$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan_inverse() +
  ggtitle("Views on Listed Medical Conditions for \nPart C EI/ECSE Services") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12),
         plot.title = element_text(color="red", size=15, hjust = 0.5),
         legend.text = element_text(size = 15))
        


```

```{r}

data_rec <- data %>%
  select(position, info_adequate_both, detail_need_both, detail_help_both, detail_covid_both) %>%
  drop_na()


pos_rec1 <- data_rec %>%
  select(position, info_adequate_both) %>%
  group_by(position, info_adequate_both) %>%
  count(info_adequate_both) %>%
  rename("response" = "info_adequate_both",
         "Adequate information\n is given ?" = "n")


pos_rec2 <- data_rec %>%
  select(position, detail_need_both) %>%
  group_by(position, detail_need_both) %>%
  count(detail_need_both) %>%
  rename("response" = "detail_need_both",
         "Detailed list\nis needed ?" = "n")
  

pos_rec3 <- data_rec %>%
  select(position, detail_help_both) %>%
  group_by(position, detail_help_both) %>%
  count(detail_help_both) %>%
  rename("response" = "detail_help_both",
         "Detailed list\n would be helpful ?" = "n")
 


pos_rec4 <- data_rec %>%
  select(position, detail_covid_both) %>%
  group_by(position, detail_covid_both) %>%
  count(detail_covid_both) %>%
  rename("response" = "detail_covid_both",
         "Detailed list during\n Covid-19 is needed ?" = "n")



data_reccomended <- left_join(pos_rec1, pos_rec2, by = c("position", "response"))
data_reccomended <- left_join(data_reccomended, pos_rec3, by = c("position", "response"))
data_reccomended <- left_join(data_reccomended, pos_rec4, by = c("position", "response"))

data_reccomended <- data_reccomended %>%
  pivot_longer(cols = -c(position, response),
               names_to = "information",
               values_to = "count") %>%
  mutate(percent = round((count/81)*100, digits = 0))

data_reccomended

# for the diverging plot

data_reccomended <- data_reccomended %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))

recommendation <- ggplot(data_reccomended,
                         aes(x = percent, y = fct_reorder(information, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.6) +
  facet_wrap(~position) +
  geom_vline(xintercept = 0, color = "#A0A0A0", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(data_reccomended$percent),
                     labels = paste0(abs(pretty(data_reccomended$percent)), "%")) +
  geom_text(aes(x = percent/2, y = information, 
                label = paste0(abs(data_reccomended$percent), "%")), color = "black", 
            size = 4) +
  #theme_xaringan() +
  theme_minimal() +

 # theme_xaringan_inverse() +
  ggtitle("(d) Recommendations/Views: All Participants") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12, color = "black"),
         plot.title = element_text(color="black", size=14, hjust = 0.5),
         legend.text = element_text(size = 14),
         strip.background = element_blank(),
         strip.text = element_text(size = 13, color = "black", hjust = 0.5)) +
  scale_fill_manual(values = c("#D0D0D0", "#B8B8B8"))

recommendation


```

### By Profession

```{r}
data_rec <- data %>%
  select(position, info_adequate_both, detail_need_both, detail_help_both, detail_covid_both) %>%
  drop_na()


pos_rec1 <- data_rec %>%
  select(position, info_adequate_both) %>%
  group_by(position, info_adequate_both) %>%
  count(info_adequate_both) %>%
  rename("response" = "info_adequate_both",
         "Adequate information\n is given ?" = "n")


pos_rec2 <- data_rec %>%
  select(position, detail_need_both) %>%
  group_by(position, detail_need_both) %>%
  count(detail_need_both) %>%
  rename("response" = "detail_need_both",
         "Detailed list\nis needed ?" = "n")
  

pos_rec3 <- data_rec %>%
  select(position, detail_help_both) %>%
  group_by(position, detail_help_both) %>%
  count(detail_help_both) %>%
  rename("response" = "detail_help_both",
         "Detailed list\n would be helpful ?" = "n")
 


pos_rec4 <- data_rec %>%
  select(position, detail_covid_both) %>%
  group_by(position, detail_covid_both) %>%
  count(detail_covid_both) %>%
  rename("response" = "detail_covid_both",
         "Detailed list during\n Covid-19 is needed ?" = "n")



data_reccomended <- left_join(pos_rec1, pos_rec2, by = c("position", "response"))
data_reccomended <- left_join(data_reccomended, pos_rec3, by = c("position", "response"))
data_reccomended <- left_join(data_reccomended, pos_rec4, by = c("position", "response"))

data_reccomended <- data_reccomended %>%
  pivot_longer(cols = -c(position, response),
               names_to = "information",
               values_to = "count") %>%
  mutate(percent = round((count/81)*100, digits = 0))

# for the diverging plot

data_reccomended <- data_reccomended %>%
  mutate(percent = ifelse(response == "Yes",
                          percent, 
                         -1*percent))

ggplot(data_reccomended,
       aes(x = percent, y = fct_reorder(information, percent), fill = response)) +
  geom_bar(aes(fill = response), stat = "identity", position = "identity", width = 0.4) +
  facet_wrap(~position) +
  geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
  colorblindr::scale_fill_OkabeIto(name = "response") +
  scale_x_continuous(breaks = pretty(data_reccomended$percent),
                     labels = paste0(abs(pretty(data_reccomended$percent)), "%")) +
  geom_text(aes(x = percent/2, y = information, label = paste0(abs(data_reccomended$percent), "%")), color = "red", 
            size = 4) +
  theme_xaringan_inverse() +
  ggtitle("Views on Listed Medical Conditions for Part C \nEI/ECSE Services by Profession") +
   theme(axis.title = element_blank(),
         panel.grid.minor = element_blank(),
         legend.title = element_blank(),
         legend.position = "bottom",
         legend.direction = "horizontal",
         plot.title.position = "plot",
         axis.text=element_text(size=12),
         plot.title = element_text(color="red", size=15, hjust = 0.5),
         legend.text = element_text(size = 15),
         strip.background = element_blank(),
         strip.text = element_text(face="bold", size = 15, color = "yellow", hjust = 0.5))



```

```{r, fig.width=10, fig.height=8}
article_plot <- plot_aware_byprofession + In_practice_all + 
  covid_eiecse + recommendation 
article_plot

```

```{r}
# save the plot in same directory.
pdf(file="article_plot.jpg")
article_plot
dev.off()
```


INTERVIEWS
=========================================

Column{data-width=20%}
-----------------------------------------------------------------------

### **Total number of interview participants**

```{r}

valueBox(18, icon = "fas fa-thumbs-up")

```

### EI/ECSE professionals

```{r}

valueBox(11)

```



### Pediatricians

```{r}

valueBox(4)

```

### Physical Therapists

```{r}

valueBox(2)

```

### Speech and Language Pathologist


```{r}

valueBox(1)

```




Row {.tabset .tabset-fade}
-------------------------------------

### Theme 1: Medical Conditions List

```{r}

grViz("digraph{

 graph[rankdir = LR]
 
 node[shape = rectangle, style = filled]
 subgraph cluster_0 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'Thistle'
 label = 'Summary'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 S1[label = 'Professionals reported that they learned about the list of medical \nconditions that make children directly eligible for Part C services, through \ntheir academic training or on-the-job. However, when we showed them \nthe medical conditions list, available on the Oregon state plan policy guidance, \nthe majority of them (90%) said, they had not seen the list before. \nFurthermore, the Oregon state medical conditions list is not referred \nto or utilized by professionals on a regular basis. Some EI agencies \nhave collated their own list for regular use in alignment with the state list.']
 }
 
 
 
 subgraph cluster_1 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'lightblue'
 label = 'Response\n(examples)'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 
 R1[label = 'We get that education as trainees in residence \nbut a lot of that i think without a real context to \nput it in. Someone may have told me about a list \nof diagnoses. I was in California where I learned about \na similar list which would automatically patients for \nEI services.']
 R2[label = 'I think it is in our handbook but you know \nlike not that list that you showed me but you know \nour version of the list']
 R3[label = 'So frankly I do not look at that list that \nyou had ever. We have an internal list \nthat we go by.(EI/ECSE) \nI don’t think I ever personally refer to the \nmedical conditions list.I would just send \nthe referral and trust the EC cares \n would figure out if it is inappropriate.(Pediatrician)']
 
 
 }
  
 subgraph cluster_2 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'SandyBrown'
 label = 'Codes'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 
 A[label = 'First time learned about the list']
 B[label = 'Awareness']
 C[label = 'Utilization']

 
 }
 
 subgraph cluster_3{
 graph[shape = rectangle]
 style = rounded
 bgcolor = DarkSeaGreen
 label = 'Theme'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 Z1[label = 'Medical Conditions List']
 
 }
 
 
 edge[color = black, arrowhead = vee, arrowsize = 1.25]
 Z1 -> {A B C }
 A -> {R1}
 B -> {R2}
 C -> {R3}
 R2 -> {S1}
 
 
 }")


```
### Theme 2: In-Practice

```{r}

grViz("digraph{


      graph[rankdir = LR]
      
      node[shape = rectangle, style = filled]
      
      subgraph cluster_0 {
      graph[shape = rectangle]
      style = rounded
      bgcolor = 'Thistle'
      label = 'Summary'
      node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
      S1[label = 'EI professionals follow standard referral and evaluation process \nwhere children with diagnosed medical conditions and physician \n statement skip eligibility determination. Healthcare professionals \nreported that they do not provide physician statement for all referrals \nwith medical condition. They also reported limited understanding on how \n having a medical diagnosis and physician statement help children, \nfamilies and EI evaluation team.  Having a medical diagnosis and physician \nstatement not only help children to receive Part C services sooner, \nbut also they are eligible for other services such as specialized \nequipment, diaper, hearing aids, augmentative communication aid, \nand social security. EI professionals also reported to have county-to-county \nvariability in the use of list of medical conditions becasue some \nhave collated their own list.']
      
      }
      
      subgraph cluster_1 {
      
      graph[shape = rectangle]
      style = rounded
      bgcolor = 'lightblue'
      label = 'Response\n(examples)'
      node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
      
      R1[label = 'To be sure I think to myself if this is a patient I know has \na high-risk of developmental delay and  sends off the referrals \nto EI program and they might tell me I am wrong. I cannot remember if \nthat has ever happened but I do trust them to some degree to \nknow who is appropriate and who is not.(Pediatrician)']
      
      R2[label = 'Sometimes if you are testing a child and then it takes a long time \nand then you also have to do an IFSP and that can \nbe too much       but for the physician statement it is usually a lot easier.']
      
      R3[label = 'Certainly having a medical diagnosis would have made it much easier \nand they would have got services earlier. Some of the things \n       are they can get specialized equipment through the state if they \nhave the medical diagnosis. They could get something as simple \nas diapers if they are of a certain age or an alternative communication device. \nThey can also get more support like personal support workers to \nsupport their families and more funding from the state if they have a \nmedical diagnosis.']
      
      R4[label = 'It is important to have a diagnosis to help them to get services \nearlier because if we do not have that then we have \nto do testing and go by scores. For small young kids who we have \nreally low expectations for their skills, being one \nor two months old, it is just really hard to get them in \nwithout having that diagnosis.']
      
      R5[label = 'I do think that there is county-to-county variability based on \nwhat I have heard from staff working in other counties.']
      
      }
      
      subgraph cluster_2{
      graph[shape = rectangle]
      style = rounded
      bgcolor = 'SandyBrown'
      label = 'Codes'
      node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
      
      A[label = 'Referral & evaluation process']
      B[label = 'Skip eval/shorten 45-day Timeline']
      C[label = 'Having med. diagnosis helps children']
      D[label = 'Refer to get physician statement']
      E[label = 'County to county variability']
     
      }
      
      
      subgraph cluster_3{
      graph[shape = rectangle]
      style = rounded
      bgcolor = DarkSeaGreen
      label = 'Theme'
      node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
      Z1[label = '    In-Practice   ']
      
      }
      edge[color = black, arrowhead = vee, arrowsize = 1.25]
      
      Z1 -> {A B C D E}
      A ->{R1}
      B ->{R2}
      C ->{R3}
      D ->{R4}
      E ->{R5}
      R3->{S1}

      
      }")
 

```
### Theme 3: During Covid-19 Pandemic

```{r}


grViz("digraph{


 graph[rankdir = LR]
 
 node[shape = rectangle, style = filled]
 
 subgraph cluster_0 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'Thistle'
 label = 'Summary'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 S1[label = 'The interview data analysis indicated that as an urgent response \nto covid-19 pandemic, the EI/ECSE and healthcare professionals \nshifted to online platform for evaluations, assessments and other \nservices. Specifically, professionals indicated that virtual platform \nis not suitable for delivering services to young children to \ncarry out evaluations and assessments. In addition, virtual platform \nhas put more burden on families to carry out observations \nwith no technical expertise and lack of suitable materials. \nSome professionals also discussed that having a well-written \nmedical conditions list and awareness among professionals \nwould have benefitted some families who had to go through \nthe eligibility determination process where \nthey could have by-passed it during the covid-19 pandemic.']
 
 }
 
 subgraph cluster_1 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'lightblue'
 label = 'Response\n(examples)'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]

 R1[label = 'Well, I know that our numbers are a lot lower \nso we are missing children. We are not getting \nthe referrals that we did pre-covid. Pretty \nsignificantly and that’s across the state of Oregon.']
 
 R2[label = 'The evaluations have been delayed because \nof covid. We were not set up for this so it \ngot delayed. Also, we have families medically \nfragile kids, they were afraid to even come \nin for the evaluation, or people just being \n nervous about the whole thing. \nA lot of families did not have a device \nor internet. We provided Ipads.']
 
 R3[label = 'Besides just the logistics of doing things \nonline, not being able to touch. I know \nour colleagues who are physical therapists \nare kind of frustrated because you can’t \nalways see the tone issue across the \nscreen. So the assessment is a little bit different \nand it very much depends on the kid.']
 
 R4[label = 'The platform sometimes gets glitchy or drops \nthe call or you know whatever or the family \ndoes not have very good video.']
 }
 
 subgraph cluster_2 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'SandyBrown'
 label = 'Codes'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 A[label = 'Number of referrals/\nFrequency of meetins']
 B[label = 'Challenges faced by families']
 C[label = 'Challenges faced by professionals']
 D[label = 'Challenges related to service delivery']
 
 }
 
 subgraph cluster_3 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = DarkSeaGreen
 label = 'Theme'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 Z1[label = 'During Covid-19 Pandemic']
 }
 
 
 edge[color = black, arrowhead = vee, arrowsize = 1.25]
 
 Z1 -> {A B C D}
 A -> {R1}
 B -> {R2}
 C -> {R3}
 D -> {R4}
 R2 -> {S1}
 
 }")



```

### Theme 4: Recommendations

```{r}

grViz("digraph{

 graph[rankdir = LR]
 
 node[shape = rectangle, style = filled]
 
 subgraph cluster_0 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'Thistle'
 label = 'Summary'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 
 S1[label = 'We investigated views of participants \non the current list and if they would \nlike to suggest any changes. Respondents \nappreciated that there is a list of medical \nconditions in Oregon, however they \ngreatly agreed in their replies that \nthe current list is not seen by them and they \ndo not refer it. The three most \nagreed recommendations are to have \na more comprehensive, detailed list, widely \ncirculated and presented in relevant forums \n on a regular basis and it should be \n embeded within the current system \n such as in the universal referral \nform and in the evaluation packet in \nEI programs to promote its appropriate utilization \nand increased understanding among \nprofessionals and families.']
 
 }
 
 
 subgraph cluster_1 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'lightblue'
 label = 'Response\n(examples)'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 R1[label = 'I am sure I have colleagues who know exactly \nhow this works. But I think more people are like \nme and not quite sure how it works and \nthen think that it probably works well about \n80% of the time.(Neonatologist)']
 R2[label = 'Having more people know about it. You know, \n it is surprising to me having worked in the field of \nearly intervention for 22 years and I have \nnever seen an actual list. (EI/ECSE Specialist)']
 R3[label = 'I have had a lot of presentations by EI folks and it has \nnever been presented to me so maybe when they go out to \ndo presentations to practices to talk about their services, \nmaybe that should be part of what they talk about.
 (Pediatrician)']
 }
 
 subgraph cluster_2 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = 'SandyBrown'
 label = 'Codes'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 A[label = 'Views on the current list']
 B[label = 'Accessible/Visible']
 C[label = 'Suggestions']
 }

 subgraph cluster_3 {
 graph[shape = rectangle]
 style = rounded
 bgcolor = DarkSeaGreen
 label = 'Theme'
 node[shape = rectangle, fillcolor = LemonChiffon, margin = 0.25]
 Z1[label = 'Recommendations']
 
 }
 
 edge[color = black, arrowhead = vee, arrowsize = 1.25]
 
 Z1 -> {A B C}
 A-> {R1}
 B -> {R2}
 C -> {R3}
 R2 -> {S1}


}")


```


Results
========================================

Row {.tabset .tabset-fade}
-------------------------------------
    
### <span style="color:  #40B0A6;">**Research Question1:**</span>

<span style="color: #E66100;">**To what extent are EI professionals aware of the categories of medical conditions that are listed in the Oregon State Plan policy guidance? (AWARENESS)**</span>

Overall analysis showed that **74%** are aware of the medical conditions listed in the Oregon State Plan Policy guidance, **72%** are aware of the medical documents and reports that constitute written documentation of eligibility, **68%** know that there are medical conditions that may not be presently adversely affecting child's development, and **56%** are aware that there are medical conditions eligible for Part C services but may not be listed in the policy guidance. When results were analyzed by profession, significant differences were found in the awareness of healthcare and EI/ECSE professionals where **21%** of healthcare professionals are aware and **19%** are not aware of the medical conditions list. Among healthcare professionals, **18%** are aware and **22%** are not aware of the written documentation of Part C eligibility, **19%** are aware and **22%** are not aware that there are medical conditions that may not be presently affecting the child's development, and only **13%** are aware and **27%** are not aware that there are medical conditions that may be eligible for Part C services but are not listed in the policy documents. No significant differences were found based on job status (full-time vs part-time), years of service, and professional role, however, a lower number of professionals **(27% vs 31%)** with less than 15 years of experience reported to be aware of medical conditions that may be eligible for EI/ECSE services but not listed. Analysis based on professional role indicated that **24%** of professionals whose role is to make referrals or provide medical diagnosis are aware of the medical conditions that may be eligible for services but not listed whereas **30%** reported being unaware.
Cronbach's alpha score of **0.73** indicates higher internal consistencies among 4 items with binary response in the survey that were asked to explore the awareness level. 

Qualitative analysis aligns with the survey response where respondents indicated that they learned about the diagnosed medical conditions that have a high probability of developmental delay, through their educational training or on-the-job. The majority of respondents (90%) said they have not seen the Oregon list of medical conditions before. EI professionals reported that programs have their own list of medical conditions. Lastly, healthcare professionals reported limited understanding of the purpose of the medical conditions list and a lack of understanding in who the list is for!  

 
### <span style="color: #40B0A6;">**Research Question 2:**</span>

<span style="color: #E66100;">**How are the categories of medical conditions that are listed in the Oregon State Plan policy guidance are understood and used by professionals in EI eligibility evaluations and assessments for birth to 5-year-old-children in Oregon? (IN PRACTICE)**</span>


Results were analyzed based on the overall response, profession, job status, years of service, and professional role. Consistent results were found across all groups indicating limited use of medical conditions list due to the lack of information. Overall analysis indicates that **77%** of professionals ask for information on medical conditions from colleague to decide on EI referrals and eligibility evaluation, **58%** have experienced a delay in EI eligibility evaluation due to lack of information, **57%** have experienced a delay in timely receipt of EI services for children with diagnosed medical conditions, and **53%** have looked for more information on medical conditions for EI eligibility determination.
**Sixty-eight percent** reported that they have not skipped EI eligibility evaluation for children with diagnosed medical conditions that have a high probability of developmental delay.

Qualitative analysis provided similar insights where healthcare professionals reproted inconsistency in providing physician statement. They also reported limited understanding on how having a medical diagnosis and physician statement help the child, family, and the EI evaluation team. EI agencies have collated their own list of medical conditions in alignment with the state list, which may lead to county-to-county variation in the referral and evaluation process. 


### <span style="color: #40B0A6;">**Research Question 3:**</span>

<span style="color: #E66100;">**How and to what extent are EI eligibility evaluation and assessments for children with diagnosed medical conditions impacted during the covid-19 restrictions? (During COVID-19)**</span>

During the COVID-19 pandemic, **94%** professionals reported to provide services remotely, **62%** experienced delay in EI eligiblity determination for children with diagnosed medical conditions, and **48%** have experienced decrease in number of children accessing healthcare and part C EI/ECSE services.
A set of questions were asked only from EI/ECSE professionals to explore the state of uptake of EI/ECSE serivces during the pandemic where **77%** reported to follow the 45-day timeline to do EI eligibility evaluation, assessments and IFSP development. **Forty percent** EI/ECSE professionals are using only tools that can be administered by parents during the COVID-19 pandemic and **30%** reported to meet children and families in-person or in home visits. 

Only **23%** EI/ECSE professionals have reported to skip EI eligibility evaluation for children referred with written documentation of eligibility, and **9%** reported to enroll children with interim IFSP where EI eligibility evaluation was not possible for children with diagnosed medical conditions during the COVID-19 pandemic. Low percentage of professionals reporting to skip EI eligibility evaluation and enrolling children with interim IFSP indicate lack of awareness or information among EI/ECSE professionals of federal regulations on children referred with diagnoed medical conditions that have a high probability of developmental delay or disability. 

The qualitative analysis confirmed the above findings. As an urgent response to covid-19 pandemic, the EI/ECSE and healthcare professionals shifted to online platform for evaluations, assessments and other services. Specifically, professionals indicated that virtual platform is not suitable for delivering services to young children to carry out evaluations and assessments. In addition, virtual platform has put more burden on families to carry out observations with no technical expertise and lack of suitable materials. Some professionals also discussed that having a well-written medical conditions list and awareness among professionals would have benefitted some families who had to go through the eligibility determination process where they could have by-passed it during the covid-19 pandemic. Results also indicated that number of referrals were severely reduced due to the covid-19 pandemic. 

### <span style="color: #40B0A6;">**Research Question 4:**</span>

<span style="color: #E66100;">**What changes do EI professionals in Oregon recommend in the level of information provided about the categories of medical conditions that are likely to result in developmental delay? (VIEWS)**</span>

The analysis resulted in a consistent response from EI/ECSE and healthcare professionals in terms of changes they would prefer in medical conditions listed in the Oregon State Plan Policy guidance. **Ninety percent** reported having a more detailed and comprehensive list would help to decide when EI eligibility evaluations can be skipped and direct referrals are made for EI assessments and **86%** percent agreed that a more detailed and comprehensive list of diagnosed medical conditions is needed in Oregon. **Seventy-nine percent** said a detailed and comprehensive list would have been helpful during the COVID-19 pandemic. **Sixty-five percent** said adequate information is not provided in the list of medical conditions that make children eligible for Part C EI/ECSE services in Oregon.

During the semi-structured interviews, we investigated views of participants on the current list and if they would like to suggest any changes. Respondents appreciated that there is a list of medical conditions in Oregon, however they greatly agreed in their replies that the current list is not seen by them and they do not refer it. The three most agreed recommendations are to have a more comprehensive, detailed list, widely circulated and presented in relevant forums on a regular basis and it should be embeded within the current system such as in the universal referral form and in the evaluation packet in EI programs to promote its appropriate utilization and increased understanding among professionals and families.


Conclusion
=========================================

Row {.tabset .tabset-fade}
-----------------------------------

### <span style="color: #40B0A6;">**Suggestion1**</span>


<span style="color: #E66100;">**Problem: According to professionals people may think if condition is not on the list, it is not eligible becasue list contains limited information.**</span>.

<span style="color: #40B0A6;">**Statement:**</span>

*“I suppose the trouble with A list is that there's always those outlying diagnoses where, ‘oh, it’s not on the list so it doesn't fit. A list of specific conditions would be nice but with the caveat of it being like this isn't necessarily an all-inclusive list.”- Speech and Language Pathologist*

*“I think more information would always be better. A lot of times we will reference this and if the condition is on there then you know it's a lot easier for us to bring the child in but I mean there's a lot of times where we have conditions that aren't on this list and so then we kind of have to go and do our own research. So if we find research that points to the child having a condition or there's been research studies that show that for the potential of developmental delay then we cite that kind of documentation and use that to bring them in if it’s not on this list.”- Physical Therapist*

*“One conditions was cleft lip and palate. So I actually don't know if that's on the list but if it’s not, certainly it should, because it has a huge developmental impact, even with mild cases.” – Speech and Language Pathologist*

<span style="color: #40B0A6;">**Recommendation:**</span>

Having more detailed, comprehensive list in simple language with more conditions that are almost 100% of the time eligible will be helpful. However, It would be helpful to have disclaimer that the list is not an all-inclusive list and that for more information a reliable web resource on medical conditions need to be provided. 

### <span style="color: #40B0A6;">**Suggestion2**</span>

<span style="color: #E66100;">**Problem: Lack of clarity in the purpose and utilization of the list? Is it for medical professionals, families or Early Interventional professionals?**</span>.

<span style="color: #40B0A6;">**Statement:**</span>

*“I think it's very important for medical professionals to be sharing this with us, the teachers who should be looking out for these things and maybe even checking in with the parents and asking them if they have concerns.”- EI/ECSE professional.*

*"I would just send the referral and trust the EC cares would figure out if it’s inappropriate."- Pediatrician*

<span style="color: #40B0A6;">**Recommendation:**</span>

Need more streamlined, clear presentation of information in simple langauge with instruction on how list can be utlized and implemented.  

### <span style="color: #40B0A6;">**Suggestion3**</span>

<span style="color: #E66100;">**Problem: Respondents varied greatly in their replies to making the list accessible for families**</span>.

<span style="color: #40B0A6;">**Statement:**</span>

*“I would love a list that I can share with a parent that they can understand without me needing to translate it alot. May be something in more layman’s terms.” (EI/ECSE professional)*

*“It’s certainly a very involved document and you need some sense of background to understand a lot of it but I think it's important to be transparent with parents about this. I would think if any child might fall into any of these things a bit near, they should be provided to parents after birth or at the point which pediatricians or Health Providers are evaluating and assessing children and helping so this should be shared.” – (EI/ECSE professional)*

*“I don't know how many parents would prefer this. I don't know if it's in super family-friendly language.” – EI/ECSE professional*

*“I don't know if they had families look through it and given feedback I think that would be a good place to start.” -Physical Therapist*

*“It is always great to have information available for families that is appropriate and relevant and in family-friendly language. However, just the idea that the process of a physician statement is so different from a standard referral so most of the time when somebody referred for a physician or a med statement because of his medical condition. The doctor is saying here's a program that will help you because we know things will go wrong. So it’s an active referral for developmental delay versus passive referral for a physician statement. So I don’t know if families would access it really all that much because they are already in the program by the time they would need it.” EI/ECSE specialist*

<span style="color: #40B0A6;">**Recommendation:**</span>

Parent groups should be consulted as part of the decision making process to get their feedback on how useful list would be for families. 


### <span style="color: #40B0A6;">**Suggestion4**</span>

<span style="color: #E66100;">**Problem: Current list is not accessible/visible/widely circulated. Only 10% interviewees had seen the Oregon list of medical conditions before the interview. Professionals were aware that children can be automatically eligible for EI services based on certain medical conditions, however, majority of participants (90%) have not seen the list before or referred in their professional role**</span>.

<span style="color: #40B0A6;">**Statement:**</span>

*“I’ve had a lot of presentations by EI folks and it’s never been presented to me so maybe when they go out to do presentations… they often go around to practices to talk about their services, maybe that should be part of what they talk about.” – Pediatrician*

*“ I don't think any parents have ever seen anything like this. I don't think parents really understand. A lot of parents don't even understand what we are, what we do.” – EI/ECSE professional*

*“Having more people know about it. You know it's surprising to me having worked in the field of early intervention for 22 years and I've never seen an actual list and if this is from the state of Oregon it seems like with all the paperwork that I do get and all the training that I do get it would be easy to be like…. Hey, here is a sheet of paper you should be aware of.”*


<span style="color: #40B0A6;">**Recommendation:**</span>

a. List should be easily available electronically and widely distributed across pediatric centers and early intervention agencies in Oregon. 

b. An electronic list with a link for medical resource for more information should be provided. 

c. List can be published on popular websites such as Oregon Pediatric Society and Family Practice Society so that healthcare professionals do not miss this. 

d. List should be included in relevant presentations and forums regularly in Oregon.


### <span style="color: #40B0A6;">**Suggestion5**</span>

<span style="color: #E66100;">**Problem: Current list is not embedded in the systems and practices in Oregon**</span>.

<span style="color: #40B0A6;">**Statement:**</span>

*“But they don’t list any on the referral form, they give examples like that they have autism but there is not the entire list there and it doesn’t refer to a list. So it could even just say look at this website, if you want to see the list of what we are asking for,“ – Pediatrician*

*“We have many children who are referred in, may be 18 months old or 2 or 3 or 3.5 years and we go ahead and then we start this standardized evaluation eligibility process. Never have I been cued to before I start to say bring this out and review something like this with the parent to make sure that the child doesn't fit into one of these categories. We don't ask the parent, especially during covid-19, we could by-pass the standardized evaluations.”- EI/ECSE specialist*

<span style="color: #40B0A6;">**Recommendation:**</span>

a. List can be made available like a reference list on the universal referral form as an additional sheet. An online version can include pop up connected to the list of medical conditions would potentially open up the opportunity for medical professionals to understand and utilize it better. 

b. List can be included in the evaluation packet in EI/ECSE programs with an instruction to review these conditions before your proceed with a standardized evaluation. 

c. List should be  included in orientation and induction of new recruits in healthcare and early intervention programs.


Column{data-height=70%}
-----------------------------------------------------------------------

### <span style="color: #40B0A6;">**Implementation Science**</span>

To have a long-term, consistent practice change for increased utilization, awareness and cross-sector uniformity (Part C Early Intervention and Healthcare) in the use of list of medical conditions for seamless identification, referral and uptake of Part C EI services for birth to 3-year-old children with diagnosed medical conditions, Oregon state need to adopt a focussed approach using the implementation science lens. 
Based on the in-depth data anlysis, we recommend use of Plan-DO-Study-Act (PDSA) process that encourages (i)planning the change(Plan), (ii) implementing the change on a small scale (Do), (iii) analyzing data and results(Study), and (iv) refining the changes based on the results(Act).

<span style="color: #E66100;">**PLAN**</span>: To set up a working committee consisting of professionals from the EI evaluation team (physical therapist, speech and langauge therapist), pediatrician/care manager, a representative of parent group, and a state representative. The purpose of this committee is to devise a new comprehensive, detailed medical conditions list with appropriate information, presented in a user-friendly language and manner. Additionally, committee will also devise a focussed dissemination and implementation plan at the planning stage. 

<span style="color: #E66100;">**DO**</span>: Embed the new list in existing processes and practices including trainings, forums, orientation and induction in a selected EI program and a set of healthcare professionals on a small scale. 

<span style="color: #E66100;">**STUDY**</span>: Collect feedback and analyze the results.

<span style="color: #E66100;">**ACT**</span>: Implement the recommended changes and go through the cycle again until intended results for sustainable use are achieved. Followed by state-wide implementation of the list of medical conditions in systems, processes, and practices.   



### <span style="color: #40B0A6;">**Conclusion**</span>

Lack of awareness, access, and inconsistent utilization of the list of diagnosed medical conditions that give children direct access to Part C EI/ECSE services in the state of Oregon indicate that a more comprehensive and detailed list in a user-friendly language is needed.  For a long-term sustainable use and seamless cross-sectoral referral and evaluation between Part C EI and healthcare professionals, the list need to be embedded in the existing systems and practices.